Radiomics-based Machine-learning Models Can Detect Pancreatic Cancer on Prediagnostic Computed Tomography Scans at a Substantial Lead Time Before Clinical Diagnosis
Open accessAbstract
Prediagnostic CT radiomics-based ML models for PDAC detection, compared against radiologists.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
11 of 22 cited sources checked · 5 supported · 0 flagged · accuracy not yet measured
- Premisecitation
Before clinical diagnosis, CTs in PDAC patients can show subtle indirect signs of cancer (3-36 months earlier).
“In patients with PDAC, subtle indirect findings suggestive of cancer (eg, pancreatic duct cutoff or dilatation, focal pancreatic atrophy) can be present on CTs at the prediagnostic stage (ie, 3–36 months before clinical diagnosis).2,7,8”
Quote in the paperSubstantial claim1 of 3 sources checked- cites [2] Singh, 2020, Computerized tomography scan in pre-diagnostic pancreatic ductal adenocarcinoma: stages of progression and potential benefits of early intervention: a retrospective study, Pancreatology source
The Introduction describes CT scans obtained 3–36 months before PDAC diagnosis, and the Results identify pancreatic duct cutoff or dilatation as a prediagnostic finding. The paper’s listed CT findings do not include focal pancreatic atrophy, so it does not support that example in the citing sentence. (LLM, conf=0.86)
Experimental — accuracy not yet measured - cites [7] Gangi, 2004, Time interval between abnormalities seen on CT and the clinical diagnosis of pancreatic cancer: retrospective review of CT scans obtained before diagnosis, Am J Roentgenol source
- cites [8] Pelaez-Luna, 2007, Resectability of presymptomatic pancreatic cancer and its relationship to onset of diabetes: a retrospective review of CT scans and fasting glucose values prior to diagnosis, Am J Gastroenterol source
- Premisecitation
Finding PDAC before cachexia worsens performance status raises the chance that eligible patients can undergo resection.
“Second, early detection before the decline in performance status from cancer-induced cachexia increases the prospect of surgical resection in eligible patients.2,5”
Quote in the paperSubstantial claim2 of 2 sources checked- cites [2] Singh, 2020, Computerized tomography scan in pre-diagnostic pancreatic ductal adenocarcinoma: stages of progression and potential benefits of early intervention: a retrospective study, Pancreatology source
- cites [5] Hart, 2019, Is screening for pancreatic cancer in high-risk individuals one step closer or a fool's errand?, Clin Gastroenterol Hepatol source
The paper’s opening section (page unavailable) says that most PDAC is diagnosed at an advanced stage, only 10–15% of patients are candidates for potentially curative treatment including surgery, and early diagnosis requires detection before symptoms. It does not mention cachexia or declining performance status as the reason earlier detection could improve the prospect of resection. (LLM, conf=0.84)
Experimental — accuracy not yet measured
- Premisecitation
People who have new-onset diabetes plus an END-PAC score of 3 or more form a group whose PDAC risk is high enough that screening might help.
“One such cohort is subjects with new-onset diabetes and a high score (≥3) on the Enriching New-Onset Diabetes for Pancreatic Cancer (END-PAC) model.4”
Quote in the paperSubstantial claim1 of 1 sources checked- cites [4] Yuan, 2020, Diabetes, weight change, and pancreatic cancer risk, JAMA Oncol source
The Discussion (page unavailable) identifies recent-onset diabetes accompanied by weight loss as a higher-risk group and says whether that group should undergo pancreatic cancer surveillance warrants prospective study. The paper does not identify an END-PAC score of ≥3 as the criterion for that group or assess screening for people meeting that cutoff. (LLM, conf=0.86)
Experimental — accuracy not yet measured
- Premisecitation
In this high-risk group, CT at diabetes onset could raise the share of PDACs that are resectable to three times the current rate.
“A computed tomography (CT) scan performed at the onset of new-onset diabetes in such a high-risk group has the potential to increase the proportion of resectable PDACs to 3 times higher than the current resectability rate.2”
Quote in the paperSubstantial claim1 of 1 sources checked- cites [2] Singh, 2020, Computerized tomography scan in pre-diagnostic pancreatic ductal adenocarcinoma: stages of progression and potential benefits of early intervention: a retrospective study, Pancreatology source
- Premisecitation
Such signs do not specifically indicate early PDAC, as control subjects can show them too.
“Second, these findings are not specific for early PDAC because they are also seen in control subjects.2,11”
Quote in the paperSubstantial claim1 of 2 sources checked- cites [2] Singh, 2020, Computerized tomography scan in pre-diagnostic pancreatic ductal adenocarcinoma: stages of progression and potential benefits of early intervention: a retrospective study, Pancreatology source
The Results section ('Parameters assessed on CT'; page not available) compares pre-diagnostic PDAC scans with matched controls and says the controls had 'negligible findings.' This allows for some findings in controls, but does not support the unqualified suggestion that these signs are generally nonspecific for early PDAC; they were much more frequent in cases. (LLM, conf=0.80)
Experimental — accuracy not yet measured - cites [11] Toshima, 2021, CT abnormalities of the pancreas associated with the subsequent diagnosis of clinical stage I pancreatic ductal adenocarcinoma more than 1 year later: a case-control study, Am J Roentgenol source
- Premisecitation
PDAC may progress quickly from imaging-invisible disease to stage IV.
“This is a major challenge because the cancer can rapidly advance from being subclinical on imaging to stage IV.2,11”
Quote in the paperSubstantial claim1 of 2 sources checked- cites [2] Singh, 2020, Computerized tomography scan in pre-diagnostic pancreatic ductal adenocarcinoma: stages of progression and potential benefits of early intervention: a retrospective study, Pancreatology source
- cites [11] Toshima, 2021, CT abnormalities of the pancreas associated with the subsequent diagnosis of clinical stage I pancreatic ductal adenocarcinoma more than 1 year later: a case-control study, Am J Roentgenol source
- Findingown data
In the test subset, the middle time from a prediagnostic scan to PDAC diagnosis was 386 days (range 97-1092).
“Median (range) time between prediagnostic CTs of the test subset and PDAC diagnosis was 386 (97–1092) days.”
Quote in the paperSubstantial claim - Findingown data
For the whole prediagnostic group, the middle time from scan to histopathological PDAC diagnosis was 398 days (range 93-1092).
“median interval between CT and PDAC diagnosis: 398 days”
- Also stated in: § Study Participants · plain
“The median (range) time interval between prediagnostic CTs and histopathological diagnosis of PDAC was 398 (93–1092) days.”
Quote in the paperSubstantial claim - Also stated in: § Study Participants · plain
- Findingown data
SVM accuracy did not differ significantly across CT vendors (Siemens 92.7%, Toshiba 92.1%, GE 88.2%, Philips 100%; P = .30).
“there was no statistically significant difference in accuracy of SVM model when the cases in the test subset were stratified by different vendors (Siemens, Toshiba, GE, and Philips were 92.7%, 92.1%, 88.2%, and 100%, respectively; P = .30).”
Quote in the paperNot a substantial claim - Findingown data
The SVM's high specificity generalized to an independent internal dataset, with specificity 92.6%.
“The high specificity of SVM was generalizable to both the independent internal (92.6%) and the National Institutes of Health dataset (96.2%).”
Quote in the paperSubstantial claim - Findingown data
The SVM classifier achieved an accuracy of 92.2% (95% CI 86.7-93.7).
“SVM had the highest sensitivity (mean; 95% confidence interval) (95.5; 85.5–100.0), specificity (90.3; 84.3–91.5), F1-score (89.5; 82.3–91.7), area under the curve (AUC) (0.98; 0.94–0.98), and accuracy (92.2%; 86.7–93.7) for classification of CTs into prediagnostic versus normal.”
Quote in the paperSubstantial claim - Findingown data
In control subjects (n=83), radiologists noted false-positive indirect PDAC findings: 7% for R4 and 18% for R5.
“Radiologists also recorded false positive indirect findings of PDAC in control subjects (n = 83) (7% R4, 18% R5).”
Quote in the paperNot a substantial claim - Findingown data
SVM performance was higher for CTs with slice thickness ≥3 mm than for those with <3 mm (98.4% vs 85.7%, P = .02).
“The performance of the SVM model for CTs with slice thicknesses ≥3 mm was higher compared with CTs with <3-mm slice thickness (98.4% and 85.7%, respectively; P = .02)”
Quote in the paperNot a substantial claim - Findingown data
The SVM's high specificity generalized to the NIH dataset, with specificity 96.2%.
“The high specificity of SVM was generalizable to both the independent internal (92.6%) and the National Institutes of Health dataset (96.2%).”
Quote in the paperNot a substantial claim - Findingown data
The SVM classifier achieved an F1-score of 89.5 (95% CI 82.3-91.7).
“SVM had the highest sensitivity (mean; 95% confidence interval) (95.5; 85.5–100.0), specificity (90.3; 84.3–91.5), F1-score (89.5; 82.3–91.7), area under the curve (AUC) (0.98; 0.94–0.98), and accuracy (92.2%; 86.7–93.7) for classification of CTs into prediagnostic versus normal.”
Quote in the paperSubstantial claim - Findingown data
The SVM classifier achieved an AUC of 0.98 (95% CI 0.94-0.98).
“SVM had the highest sensitivity (mean; 95% confidence interval) (95.5; 85.5–100.0), specificity (90.3; 84.3–91.5), F1-score (89.5; 82.3–91.7), area under the curve (AUC) (0.98; 0.94–0.98), and accuracy (92.2%; 86.7–93.7) for classification of CTs into prediagnostic versus normal.”
Quote in the paperSubstantial claim - Findingown data
The SVM classifier achieved a specificity of 90.3% (95% CI 84.3-91.5) for distinguishing prediagnostic from normal CTs.
“SVM had the highest sensitivity (mean; 95% confidence interval) (95.5; 85.5–100.0), specificity (90.3; 84.3–91.5), F1-score (89.5; 82.3–91.7), area under the curve (AUC) (0.98; 0.94–0.98), and accuracy (92.2%; 86.7–93.7) for classification of CTs into prediagnostic versus normal.”
Quote in the paperSubstantial claim - Findingown data
The support vector machine model achieved the highest area under the curve (0.98; 95% CI, 0.94-0.98) on the test subset.
“The SVM model had the highest AUC (0.98; 95% CI, 0.94–0.98) on the test subset”
Quote in the paperSubstantial claim - Findingown data
The SVM classifier achieved a sensitivity of 95.5% (95% CI 85.5-100.0) for distinguishing prediagnostic from normal CTs.
“SVM had the highest sensitivity (mean; 95% confidence interval) (95.5; 85.5–100.0), specificity (90.3; 84.3–91.5), F1-score (89.5; 82.3–91.7), area under the curve (AUC) (0.98; 0.94–0.98), and accuracy (92.2%; 86.7–93.7) for classification of CTs into prediagnostic versus normal.”
Quote in the paperSubstantial claim - Findingown data
Interreader agreement between radiologists was only fair, with Cohen's kappa 0.3.
“In contrast, interreader radiologist agreement was only fair (Cohen’s kappa 0.3) and their mean AUC (0.66; 0.46–0.86) was lower than each of the 4 ML models (AUCs: 0.95–0.98) (P < .001).”
Quote in the paperNot a substantial claim - Findingown data
The KNN, RF, and XGBoost models had comparable AUCs of 0.95, 0.95, and 0.96, respectively.
“All 3 other ML models, KNN, RF, and XGBoost, had comparable AUCs (0.95, 0.95, and 0.96, respectively).”
Quote in the paperNot a substantial claim - Study design and conductown data
Every prediagnostic scan had already been read as showing no PDAC in routine practice.
“All these CTs had been previously interpreted to be negative for PDAC during routine clinical interpretation at our institution.”
Quote in the paperSubstantial claim - Premisecitation
Screening by risk with CT in these groups is economically valuable to society under various analyses.
“A risk-based screening strategy using CT in such cohorts has societal economic value across a range of analyses.1”
Quote in the paperSubstantial claim0 of 1 sources checked- cites [1] Schwartz, 2021, Potential cost-effectiveness of risk-based pancreatic cancer screening in patients with new-onset diabetes, J Natl Compr Canc Netw source
- Premisecitation
The NIH-PCT public set contains 80 abdominal CTs from healthy people with no pancreatic disease and morphologically normal glands.
“This public dataset consists of 80 abdomen CTs acquired in portal venous (53 men, 27 women; mean [standard deviation (SD)] age: 46.8 [16.7] years) from healthy subjects without pancreas pathology.16 All CTs have morphologically normal pancreas.”
Quote in the paperSubstantial claim0 of 1 sources checked- cites [16] Suman, 2021, Quality gaps in public pancreas imaging datasets: Implications & challenges for AI applications, Pancreatology source
- Interpretationown data
The machine-learning models detected the imaging signature of PDAC at a substantial lead time (median 398 days; range 93-1092) before clinical diagnosis.
“ML models based on first-order intensity and second-order texture features from volumetrically segmented normal pancreas detected the imaging signature of PDAC at a substantial lead time (median [range] 398 [93–1092] days) before its clinical diagnosis.”
Quote in the paperSubstantial claim - Interpretationown data
Radiomics-based machine-learning classifiers can detect imaging signatures of early PDAC from morphologically normal pancreas.
“Thus, radiomics-based ML classifiers can detect the imaging signatures of early PDAC from morphologically normal pancreas.”
Quote in the paperSubstantial claim
- Classified as probably not a substantial claim. 19
- May combine two claims. 17
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationThe Early Detection Initiative (NCT04662879) is an ongoing large prospective study of a screening strategy that uses END-PAC and CT in 12,500 people.
“A large prospective trial, the Early Detection Initiative (EDI) (NCT04662879), sponsored by the Pancreatic Cancer Action Network, is under way to evaluate outcomes of a screening strategy in 12,500 participants by using the END-PAC model and CT.6”
Quote in the paperNot a substantial claim1 of 1 sources checked- cites [6] Chari, 2021, Early detection initiative: a randomized controlled trial of algorithm-based screening in patients with new onset hyperglycemia and diabetes for early detection of pancreatic ductal adenocarcinoma, Contemp Clin Trials source
Classified as probably not a substantial claim.
Study design and conductcitationTo address possible confounding by slice thickness, features whose AUC for slice-thickness group exceeded 0.8 were considered biased and dropped.
“To reduce the potentially confounding effect of slice thickness on the ML models, we used a recently described process21 to identify and then account for potential dependence of the extracted features on slice thickness”
- Also stated in: § Exclusion of CT Slice Thickness Dependent Radiomics Features · plain
“Features with a high AUC (> 0.8) were deemed to be biased by CT slice thickness and were, therefore, removed.”
Quote in the paperSubstantial claim1 of 1 sources checked- cites [21] Lu, 2021, Uncontrolled confounders may lead to false or overvalued radiomics signature: a proof of concept using survival analysis in a multicenter cohort of kidney cancer, Front Oncol source
- Also stated in: § Exclusion of CT Slice Thickness Dependent Radiomics Features · plain
May combine two claims.
PremisecitationDetecting PDAC earlier improves survival beyond the effect of lead time, since smaller tumors are more often resectable.
“Early detection offers survival benefit beyond lead time, as tumors are smaller in volume and more likely to be resectable when detected earlier.2-4”
Quote in the paperSubstantial claim2 of 3 sources checked- cites [2] Singh, 2020, Computerized tomography scan in pre-diagnostic pancreatic ductal adenocarcinoma: stages of progression and potential benefits of early intervention: a retrospective study, Pancreatology source
The Discussion reports that earlier-detected tumors were smaller and more likely to be resectable, and says the study suggests a survival benefit beyond lead time. The Abstract Conclusion calls that benefit likely, while the Introduction says a definitive answer requires a prospective randomized trial. The claim presents the survival benefit as established rather than suggested. (LLM, conf=0.85)
Experimental — accuracy not yet measured - cites [3] Vasen, 2016, Benefit of surveillance for pancreatic cancer in high-risk individuals: outcome of long-term prospective follow-up studies from three european expert centers, J Clin Oncol source
- cites [4] Yuan, 2020, Diabetes, weight change, and pancreatic cancer risk, JAMA Oncol source
The Introduction reports earlier-stage diagnosis and longer survival among patients undergoing surveillance, and the Discussion says early-stage pancreatic cancer can be cured with aggressive treatment. The paper also describes a window in which some tumors are surgically resectable. It does not establish that the survival benefit exceeds lead-time effects or attribute greater resectability to smaller tumor volume. Page numbers are unavailable in our copy. (LLM, conf=0.84)
Experimental — accuracy not yet measured
May combine two claims.
Findingown dataIn external validation, SVM accuracy was 95% for CTs from our institution and 90.4% for CTs from external institutions.
“Of the 45 CTs from our institution, the model correctly classified 16 of 17 (94%) prediagnostic CTs and 27 of 28 (96%) control CTs, yielding an overall accuracy of 95%. Of the 83 CTs from external institutions, the model correctly classified 27 of 28 (96%) prediagnostic CTs and 48 of 55 (87%) control CTs, yielding an overall accuracy of 90.4%”
Quote in the paperSubstantial claimMay combine two claims.
Findingown dataThe support vector machine correctly classified 43 of 45 prediagnostic CTs and 75 of 83 control CTs, with overall accuracy 92.2%.
“It correctly classified 43 of 45 (95%) prediagnostic CTs and 75 of 83 (90.3%) control CTs, yielding an overall accuracy of 92.2%.”
Quote in the paperSubstantial claimMay combine two claims.
Findingown dataThe two radiologist readers had a mean classification AUC of 0.66 (CI, 0.46-0.86), significantly lower than the four ML classifiers (AUCs 0.95-0.98; P < .001).
“The mean classification AUC of the 2 readers was 0.66 (CI, 0.46–0.86), which was significantly lower than each of the 4 ML classifiers (AUCs: 0.95–0.98) (P < .001)”
Quote in the paperSubstantial claimMay combine two claims.
Findingown dataRadiologists' mean AUC was 0.66 (95% CI 0.46-0.86), lower than each of the four ML models (AUCs 0.95-0.98), with P < .001.
“In contrast, interreader radiologist agreement was only fair (Cohen’s kappa 0.3) and their mean AUC (0.66; 0.46–0.86) was lower than each of the 4 ML models (AUCs: 0.95–0.98) (P < .001).”
Quote in the paperSubstantial claimMay combine two claims.
Findingown dataThe KNN, RF, and XGB models had AUCs of 0.95, 0.95, and 0.96, comparable to the SVM's 0.98.
“Other 3 models (KNN, RF, and XGB) had comparable AUCs (0.95, 0.95, and 0.96, respectively) to SVM (0.98)”
Quote in the paperNot a substantial claimMay combine two claims.
Findingown dataSVM specificity was 93% on an independent internal set of 176 healthy subjects and 96% on the public NIH-PCT dataset (n = 80).
“of the 176 CTs of healthy subjects from the independent internal set, the SVM model correctly classified the pancreas as normal in 163 CTs, yielding a specificity of 93%. The specificity was also generalizable on the public NIH-PCT dataset (n = 80) where the SVM model correctly classified the pancreas as normal in 77 CTs, yielding a specificity of 96%.”
Quote in the paperSubstantial claimClassified as probably not a substantial claim.
Study design and conductown dataEach model's hyperparameters were tuned by randomized search with stratified fivefold cross-validation, choosing the settings with the best cross-validated AUC.
“A 5-fold stratified cross-validation–based randomized search was applied to evaluate each model’s hyperparameters. The parameters that yielded the highest cross-validation AUC were selected for each classifier model.”
Quote in the paperSubstantial claimClassified as probably not a substantial claim.
Study design and conductown dataRadiologists reviewed control scans to verify adequate quality, portal venous enhancement, and normal pancreatic morphology.
“These control CTs were also curated by radiologists to confirm optimal image quality, portal venous phase, and normal pancreas morphology.”
Quote in the paperSubstantial claimClassified as probably not a substantial claim. May combine two claims.
Study design and conductown dataOne of three radiologists rechecked every prediagnostic scan and verified adequate image quality, portal venous enhancement, no pancreatitis, no focal solid or cystic lesion, and no biliary or pancreatic stent.
“Each CT was rereviewed by 1 of 3 radiologists (R1, R2, R3 with 2–4 years of post-radiology residency experience) who confirmed optimal image quality (eg, lack of motion artifacts compromising visual interpretation), portal venous phase of enhancement, absence of pancreatitis, focal solid or cystic lesion, and biliary or pancreatic duct stent (Figure 2).”
Quote in the paperSubstantial claimMay combine two claims.
Study design and conductown dataControls were age-matched and randomly selected, and none developed PDAC over three or more years of follow-up.
“From this cohort, we selected an age-matched set of control patients (n = 265) (140 men, 125 women; median [range] age: 67 [30–89] years) who were not diagnosed subsequently with PDAC during 3 years or more of follow-up (Figure 1).”
Quote in the paperSubstantial claimClassified as probably not a substantial claim.
Study design and conductreasoningThe model that achieved the top AUC in the internal test set was then applied to the NIH-PCT dataset to check specificity externally.
“Therefore, to externally validate its specificity, the classifier with the highest AUC on the internal test subset was tested on the National Institutes of Health-Pancreas CT (NIH-PCT) dataset.”
Quote in the paperSubstantial claimMay combine two claims.
Interpretationown dataMachine-learning models using radiomics can identify PDAC in otherwise normal pancreas at a long lead time before clinical diagnosis, beyond what human readers can interrogate.
“Radiomics-based ML models can detect PDAC from normal pancreas when it is beyond human interrogation capability at a substantial lead time before clinical diagnosis.”
Quote in the paperSubstantial claimProbably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.
Study design and conductnone givenReadings took place in the same lighting and on calibrated DICOM monitors at a PACS workstation.
“Interpretation was performed under identical ambient conditions with calibrated DICOM–compliant monitors at a PACS workstation (Visage Imaging, San Diego, CA).”
Quote in the paperNot a substantial claimClassified as probably not a substantial claim. May combine two claims.
Study design and conductreasoningLASSO logistic regression was used to select features; by zeroing some coefficients it simplifies the model, aiming to limit overfitting and aid generalization.
“Feature selection was done with least absolute shrinkage and selection operator (LASSO)20 logistic regression, which is an L1-regularization–based method that penalizes the L1-norm of the feature weight coefficients. Therefore, it reduces the model’s complexity by eliminating some of the coefficients (ie, coefficients become zero) and the corresponding features to minimize overfitting and improve”
Quote in the paperSubstantial claim0 of 1 sources checked- cites [20] Tibshirani, 1997, The lasso method for variable selection in the Cox model, Stat Med source
May combine two claims.
PremisecitationClinicians frequently miss these indirect signs, which delays diagnosis.
“However, these findings are often overlooked in clinical practice, leading to delayed diagnosis.9,10”
Quote in the paperSubstantial claim0 of 2 sources checked- cites [9] Kang, 2021, The implications of missed or misinterpreted cases of pancreatic ductal adenocarcinoma on imaging: a multi-centered population-based study, Eur Radiol source
- cites [10] Kang, 2021, Factors associated with missed and misinterpreted cases of pancreatic ductal adenocarcinoma, Eur Radiol source
Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.
Study design and conductown dataThe public set provides pancreas masks; the authors checked them for correct boundaries.
“The public dataset includes pancreas segmentation masks, which we reviewed to ensure the accuracy of segmentation boundaries.”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S3 (load-bearing).
Study design and conductown dataR1, R2, and R3 jointly examined every misclassified scan to look for reasons for the error.
“All the incorrectly classified CTs were reviewed in consensus by R1, R2, and R3 to identify any potential causes for misclassification.”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S3 (load-bearing).
Study design and conductown dataFor external validation, the test subset was stratified by institution: 45 (35.2%) from our institution and 83 (64.8%) from external institutions.
“For external validation of the model with highest accuracy, we stratified our test subset according to the institution where CTs had been done: 45 (35.2%) were from our institution (17 [37.8%] prediagnostic CTs and 28 [62.2%] control CTs), whereas 83 (64.8%) were from external institutions (28 [33.7%] prediagnostic CTs and 55 [66.3%] control CTs).”
Quote in the paperNot a substantial claimClassified as probably not a substantial claim. May combine two claims.
InterpretationreasoningIf validated prospectively and combined with fluid biomarkers, these models could enable PDAC detection when surgery can still cure.
“Prospective validation and integration of such models with complementary fluid-based biomarkers has the potential for PDAC detection at a stage when surgical cure is a possibility.”
Quote in the paperSubstantial claimProbably not a substantial claim: may not pass S3 (load-bearing).
Study design and conductcitationFor scans done at the authors' institution, acquisition and reconstruction followed their earlier published protocols.
“Standard acquisition and reconstruction parameters as per our previously published protocols17 had been followed for all CTs performed at our institution.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [17] Panda, 2021, Borderline resectable and locally advanced pancreatic cancer: FDG PET/MRI and CT tumor metrics for assessment of pathologic response to neoadjuvant therapy and prediction of survival, Am J Roentgenol source
Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.
Study design and conductown dataSpecificity was also tested in a separate internal set of 176 scans and in the public NIH-PCT set of 80 normal-pancreas scans.
“Its specificity was also evaluated on an independent internal validation set (n = 176) and on the public NIH-PCT dataset (n = 80) with normal pancreas.”
Quote in the paperNot a substantial claimClassified as probably not a substantial claim.
Study design and conductown dataThe scans were split at random into a training-validation set of 292 and a test set of 128, with a 70/30 ratio.
“The preceding dataset of prediagnostic CTs (n = 155) and age-matched control CTs with normal pancreas (n = 265) was randomly divided into training-validation (n = 292) and test (n = 128) subsets using a 70% / 30% split for ML model development and testing.”
Quote in the paperSubstantial claimProbably not a substantial claim: may not pass S3 (load-bearing).
Study design and conductown dataThe most accurate model was externally validated by splitting the test set by institution.
“As mentioned previously, external validation of the model with highest accuracy was performed by stratifying the test subset according to the institution (35.2% CTs from our institution and 64.8% CTs from external institutions).”
Quote in the paperNot a substantial claimClassified as probably not a substantial claim.
Study design and conductown dataSpecificity was also checked in a separate internal set of 176 scans from people with normal pancreas who remained PDAC-free for three or more years.
“For additional independent validation of the specificity, we created another internal hold-out set of 176 CTs (68 men, 108 women; median [range] age: 45 [19–94] years) who had normal pancreas and did not develop PDAC during at least 3 years or more of follow-up.”
Quote in the paperSubstantial claimProbably not a substantial claim: may not pass S3 (load-bearing).
Study design and conductown dataThe LASSO tuning parameter was chosen by a grid search with stratified fivefold cross-validation on the training set, taking the value with the best cross-validated AUC.
“Hyperparameter for LASSO was evaluated using stratified 5-fold cross-validation–based grid search method on the training set (n = 292). The parameter that provided the highest cross-validation AUC was selected.”
Quote in the paperNot a substantial claimClassified as probably not a substantial claim.
Study design and conductown dataR4 and R5, unaware of outcomes and case mix, independently assessed the pancreas on test-set scans.
“Two radiologist readers (R4 with 9 years and R5 with 3 years of post-residency experience) who were blinded to the outcomes and the distribution of prediagnostic vs control CTs independently evaluated the pancreas in the test subset CTs.”
Quote in the paperSubstantial claimMay combine two claims.
Interpretationown dataRadiologist readers had low interreader agreement, sensitivity, and discrimination accuracy, indicating that AI-based approaches can detect PDAC at a subclinical stage beyond human interrogation.
“In contrast, radiologist readers had low interreader agreement, sensitivity, and discrimination accuracy, which shows that novel AI-based approaches can detect PDAC at a subclinical stage when it is beyond the scope of the human interrogation.”
Quote in the paperSubstantial claim
The paper relies on this without support.
In many PDAC patients, the pancreas appears morphologically normal on CT before diagnosis.
“Importantly, the pancreas tends to be morphologically normal on CT at prediagnostic stage in many patients.”
The paper relies on this without support.
No public dataset of prediagnostic CT scans exists.
“There is no public dataset of prediagnostic CTs.”
The paper relies on this without support.
The prediagnostic CT dataset used in this study is the largest reported in the literature.
“The prediagnostic CT dataset used in this study is the largest reported in the literature.”
The paper relies on this without support.
There are no public datasets of prediagnostic CTs, which is a major barrier for investigating CT for early PDAC detection.
“However, there are no public datasets of prediagnostic CTs, which is one major barrier for investigation of CT for early detection of PDAC.”
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
Mukherjee and colleagues found that SVM outperformed RF, KNN, and XGBoost in diagnostic accuracy for early detection of pancreatic carcinoma using pre-diagnostic contrast-enhanced CT scans.
“Mukherjee and colleagues compared four ML classifiers on 34 radiomics features selected from pre-diagnostic contrast-enhanced CT scans for the early detection of pancreatic carcinoma and found that SVM outperformed RF, a KNN, and XGBoost in diagnostic accuracy [14].”
From Applicability of Radiomics for Differentiation of Pancreatic Adenocarcinoma from Healthy Tissue of Pancreas by Using Magnetic Resonance Imaging and Machine Learning (2025) · cites this paper as [14]
The Study Participants section identifies the prediagnostic scans as contrast-enhanced CTs. The Results (Radiomics-based ML Models; page not available) report that SVM had the highest accuracy among the four classifiers. However, although 34 radiomic features were initially selected, two were removed before the classifiers were built, so the models used 32 features, not 34. (LLM, conf=0.92)
Experimental — accuracy not yet measuredIn our statistics, several studies focused on PC or PC precursor lesions diagnosis or prediction by AI-assisted CT, with AUC 0.79-0.999, accuracy 77.66%-99.2%, sensitivity 76.64%-100%, specificity 85.59%-98.5%.
“In our statistics, several studies focused on PC or PC precursor lesions diagnosis or prediction by AI-assisted CT. Their AUC, accuracy, sensitivity, specificity were 0.79-0.999, 77.66%-99.2%, 76.64%-100%, 85.59%-98.5%, respectively 113-122.”
From Artificial intelligence in pancreatic cancer (2022) · cites this paper as [121]
Cites
DeLong, E. R., DeLong, D. M., & Clarke-Pearson, D. L. (1988) · Biometrics · doi:10.2307/2531595 · Cited by 22,649 · see citing works
Computational Radiomics System to Decode the Radiographic Phenotype
(2017) · doi:10.1158/0008-5472.can-17-0339 · Cited by 6,593 · see citing works
Vasen, H., Ibrahim, I., Ponce, C. G., Slater, E. P., Matthäi, E., Carrato, A., Earl, J., Robbers, K., van Mil, A. M., Potjer, T., et al. (2016) · Journal of Clinical Oncology · doi:10.1200/jco.2015.64.0730 · Cited by 344 · see citing works
Role of Immune Cells and Immune-Based Therapies in Pancreatitis and Pancreatic Ductal Adenocarcinoma
Zheng, L., Xue, J., Jaffee, E. M., & Habtezion, A. (2013) · Gastroenterology · doi:10.1053/j.gastro.2012.12.042 · Cited by 323 · see citing works
Gangi, S., Fletcher, J. G., Nathan, M. A., Christensen, J. A., Harmsen, W. S., Crownhart, B. S., & Chari, S. T. (2004) · American Journal of Roentgenology · doi:10.2214/ajr.182.4.1820897 · Cited by 249 · see citing works
Self-citationChu, L. C., Park, S., Kawamoto, S., Fouladi, D. F., Shayesteh, S., Zinreich, E. S., Graves, J. S., Horton, K. M., Hruban, R. H., Yuille, A. L., et al. (2019) · American Journal of Roentgenology · doi:10.2214/ajr.18.20901 · Cited by 193 · see citing works
Pelaez-Luna, M., Takahashi, N., Fletcher, J. G., & Chari, S. T. (2007) · The American Journal of Gastroenterology · doi:10.1111/j.1572-0241.2007.01480.x · Cited by 180 · see citing works
Self-citationUnderstanding Sources of Variation to Improve the Reproducibility of Radiomics
Zhao, B. (2021) · Frontiers in Oncology · doi:10.3389/fonc.2021.633176 · Cited by 158 · see citing works
Artificial Intelligence and Early Detection of Pancreatic Cancer: 2020 Summative Review
Kenner, B., Chari, S. T., Kelsen, D., Klimstra, D. S., Pandol, S. J., Rosenthal, M., Rustgi, A. K., Taylor, J. A., Yala, A., Abul-Husn, N., et al. (2021) · Pancreas · doi:10.1097/mpa.0000000000001762 · Cited by 158 · see citing works
Self-citationA comparison of denominator degrees of freedom methods for multiple observer ROC analysis
Hillis, S. L. (2006) · Statistics in Medicine · doi:10.1002/sim.2532 · Cited by 151 · see citing works
Diabetes, weight change, and pancreatic cancer risk
Yuan, C., Babic, A., Khalaf, N., Nowak, J. A., Brais, L. K., Rubinson, D. A., Ng, K., Aguirre, A. J., Pandharipande, P. V., Fuchs, C. S., et al. (2020) · JAMA Oncology · doi:10.1001/jamaoncol.2020.2948 · Cited by 150 · see citing works
CT texture analysis of pancreatic cancer
Sandrasegaran, K., Lin, Y., Asare-Sawiri, M., Taiyini, T., & Tann, M. A. (2018) · European Radiology · doi:10.1007/s00330-018-5662-1 · Cited by 128 · see citing works
Multimodal Radiomic Features for the Predicting Gleason Score of Prostate Cancer
Chaddad, A., Kucharczyk, M. J., & Niazi, T. M. (2018) · Cancers · doi:10.3390/cancers10080249 · Cited by 110 · see citing works
Factors associated with missed and misinterpreted cases of pancreatic ductal adenocarcinoma
Kang, J. D., Clarke, S. E., & Costa, A. F. (2021) · European Radiology · doi:10.1007/s00330-020-07307-5 · Cited by 84 · see citing works
Toshima, F., Watanabe, R., Inoue, D., Yoneda, N., Yamamoto, T., Sasahira, N., Sasaki, T., Matsuyama, M., Minehiro, K., Tateishi, U., et al. (2021) · American Journal of Roentgenology · doi:10.2214/ajr.21.26014 · Cited by 81 · see citing works
Panda, A., Garg, I., Truty, M. J., Kline, T. L., Johnson, M. P., Ehman, E. C., Suman, G., Anaam, D. A., Kemp, B. J., Johnson, G. B., et al. (2021) · American Journal of Roentgenology · doi:10.2214/ajr.20.24567 · Cited by 71 · see citing works
Self-citationRigiroli, F., Hoye, J., Lerebours, R. C., Lafata, K. J., Li, C., Meyer, M., Lyu, P., Ding, Y., Schwartz, F. R., Mettu, N. B., et al. (2021) · Radiology · doi:10.1148/radiol.2021210699 · Cited by 64 · see citing works
Singh, D. P., Sheedy, S., Goenka, A. H., Wells, M., Lee, N. J., Barlow, J., Sharma, A., Kandlakunta, H., Chandra, S., Garg, S. K., et al. (2020) · Pancreatology · doi:10.1016/j.pan.2020.07.410 · Cited by 62 · see citing works
Self-citationChari, S. T., Maitra, A., Matrisian, L. M., Shrader, E. E., Wu, B. U., Kambadakone, A., Zhao, Y.-Q., Kenner, B., Rinaudo, J. A. S., Srivastava, S., et al. (2022) · Contemporary Clinical Trials · doi:10.1016/j.cct.2021.106659 · Cited by 62 · see citing works
Self-citationLi, J., Lu, J., Liang, P., Li, A., Hu, Y., Shen, Y., Hu, D., & Li, Z. (2018) · Cancer Medicine · doi:10.1002/cam4.1746 · Cited by 58 · see citing works
Schwartz, N. R. M., Matrisian, L. M., Shrader, E. E., Feng, Z., Chari, S., & Roth, J. A. (2022) · Journal of the National Comprehensive Cancer Network · doi:10.6004/jnccn.2020.7798 · Cited by 43 · see citing works
Self-citationNon-Invasive Biomarkers for Earlier Detection of Pancreatic Cancer—A Comprehensive Review
Brezgyte, G., Shah, V. M., Jach, D., & Crnogorac‐Jurcevic, T. (2021) · Cancers · doi:10.3390/cancers13112722 · Cited by 43 · see citing works
Panda, A., Korfiatis, P., Suman, G., Garg, S. K., Polley, E. C., Singh, D. P., Chari, S. T., & Goenka, A. H. (2021) · Medical Physics · doi:10.1002/mp.14782 · Cited by 41 · see citing works
Self-citationQuality gaps in public pancreas imaging datasets: Implications & challenges for AI applications
Suman, G., Patra, A., Korfiatis, P., Majumder, S., Chari, S. T., Truty, M. J., Fletcher, J. G., & Goenka, A. H. (2021) · Pancreatology · doi:10.1016/j.pan.2021.03.016 · Cited by 41 · see citing works
Self-citationNew Methodological Tools for Multiple-Reader ROC Studies
Obuchowski, N. A. (2007) · Radiology · doi:10.1148/radiol.2432060387 · Cited by 37 · see citing works
Narang, S., Kim, D., Aithala, S. P., Heimberger, A. B., Ahmed, S., Rao, D., Rao, G., & Rao, A. U. K. (2017) · Oncotarget · doi:10.18632/oncotarget.20643 · Cited by 35 · see citing works
Is screening for pancreatic cancer in high-risk individuals one step closer or a fool's errand?
Hart, P. A., & Chari, S. T. (2019) · Clinical Gastroenterology and Hepatology · doi:10.1016/j.cgh.2018.09.024 · Cited by 32 · see citing works
Self-citationLu, L., Ahmed, F. S., Akin, O., Luk, L., Guo, X., Yang, H., Yoon, J., Hakimi, A. A., Schwartz, L. H., & Zhao, B. (2021) · Frontiers in Oncology · doi:10.3389/fonc.2021.638185 · Cited by 31 · see citing works
Kang, J., Clarke, S. E., Abdolell, M., Ramjeesingh, R., Payne, J., & Costa, A. F. (2021) · European Radiology · doi:10.1007/s00330-020-07120-0 · Cited by 22 · see citing works
Suman, G., Panda, A., Korfiatis, P. D., Edwards, M. E., Garg, S. K., Blezek, D. J., Chari, S. T., & Goenka, A. H. (2020) · Abdominal Radiology · doi:10.1007/s00261-020-02741-x · Cited by 21 · see citing works
Self-citationPrognostic Value of Transfer Learning Based Features in Resectable Pancreatic Ductal Adenocarcinoma
Zhang, Y., Lobo-Mueller, E. M., Karanicolas, P. J., Gallinger, S., Haider, M. A., & Khalvati, F. (2020) · Frontiers in Artificial Intelligence · doi:10.3389/frai.2020.550890 · Cited by 16 · see citing works
Tibshirani (1997) · doi:10.1002/(sici)1097-0258(19970228)16:4
Flagged first, then by citations
3 of the 36 references couldn’t be matched to a known paper.
Cited by
Sorted by citations
Pancreatic Cancer Surveillance and Survival of High-Risk Individuals
Blackford, A. L., Canto, M. I., Dbouk, M. H., Hruban, R. H., Katona, B. W., Chak, A., Brand, R. E., Syngal, S., Farrell, J. J., Kastrinos, F., et al. (2024) · JAMA Oncology · doi:10.1001/jamaoncol.2024.1930 · Cited by 178 · see citing works
Artificial intelligence in pancreatic cancer
Huang, B., Huang, H., Zhang, S., Zhang, D., Shi, Q., Liu, J., & Guo, J. (2022) · Theranostics · doi:10.7150/thno.77949 · Cited by 142 · see citing works
“[121] 420 cases CE-CT KNN, SVM, RF, and XGBoost PDAC detection AUC (0.”
MethodWang, W., Liang, H., Zhang, Z., Xu, C., Wei, D., Li, W., Ye, Q., Zhang, L., Liu, J., & Lei, D. (2024) · EClinicalMedicine · doi:10.1016/j.eclinm.2023.102385 · Cited by 106 · see citing works
Klatte, D. C. F., Boekestijn, B., Onnekink, A. M., Dekker, F. W., van der Geest, L. G., Wasser, M. N. J. M., Feshtali, S., Mieog, J. S. D., Luelmo, S. A. C., Morreau, H. M., et al. (2023) · Gastroenterology · doi:10.1053/j.gastro.2023.02.032 · Cited by 74 · see citing works
Korfiatis, P. D., Suman, G., Patnam, N. G., Trivedi, K. H., Karbhari, A., Mukherjee, S., Cook, C. J., Klug, J. R., Patra, A., Khasawneh, H., et al. (2023) · Gastroenterology · doi:10.1053/j.gastro.2023.08.034 · Cited by 79 · see citing works
Self-citationFrom Machine Learning to Patient Outcomes: A Comprehensive Review of AI in Pancreatic Cancer
Tripathi, S., Tabari, A., Mansur, A., Dabbara, H., Bridge, C. P., & Daye, D. (2024) · Diagnostics · doi:10.3390/diagnostics14020174 · Cited by 64 · see citing works
Yao, L., Zhang, Z., Keles, E., Yazici, C., Tirkes, T., & Bagci, U. (2023) · Current Opinion in Gastroenterology · doi:10.1097/mog.0000000000000966 · Cited by 53 · see citing works
Tumor immune microenvironment in pancreatic ductal adenocarcinoma revisited – Exploring the “Space”
Bräutigam, K., Skok, K., Szymoński, K., Rift, C. V., & Karamitopoulou, E. (2025) · Cancer Letters · doi:10.1016/j.canlet.2025.217699 · Cited by 57 · see citing works
Mukund, A., Afridi, M. A., Karolak, A. M., Park, M. A., Permuth, J. B., & Rasool, G. (2024) · Cancers · doi:10.3390/cancers16122240 · Cited by 50 · see citing works
Rai, H. M., Yoo, J., & Razaque, A. (2024) · Expert Systems with Applications · doi:10.1016/j.eswa.2024.124838 · Cited by 54 · see citing works
Artificial Intelligence in Pancreatic Imaging: A Systematic Review
Podină, N., Gheorghe, E. C., Constantin, A., Cazacu, I., Croitoru, V., Gheorghe, C., Balaban, D. V., Jinga, M., Țieranu, C. G., & Săftoiu, A. (2025) · United European Gastroenterology Journal · doi:10.1002/ueg2.12723 · Cited by 48 · see citing works
Søreide, K., Ismail, W., Roalsø, M. T. T., Ghotbi, J., & Zaharia, C. (2023) · Cancer Control · doi:10.1177/10732748231154711 · Cited by 47 · see citing works
A primer on artificial intelligence in pancreatic imaging
Ahmed, T. M., Kawamoto, S., Hruban, R. H., Fishman, E. K., Soyer, P. A., & Chu, L. C. H. (2023) · Diagnostic and Interventional Imaging · doi:10.1016/j.diii.2023.03.002 · Cited by 43 · see citing works
Role of Machine Learning in Precision Oncology: Applications in Gastrointestinal Cancers
Tabari, A., Chan, S. M., Omar, O. M., Iqbal, S. I., Gee, M. S., & Daye, D. (2022) · Cancers · doi:10.3390/cancers15010063 · Cited by 37 · see citing works
Barat, M., Pellat, A., Hoeffel, C. C., Dohan, A. J. J., Coriat, R., Fishman, E. K., Nougaret, S., Chu, L. C. H., & Soyer, P. A. (2023) · Japanese Journal of Radiology · doi:10.1007/s11604-023-01504-0 · Cited by 35 · see citing works
Majumder, S., Katz, S., Kontos, D., & Roshkovan, L. (2023) · BJR|Open · doi:10.1093/bjro/tzad004 · Cited by 42 · see citing works
Self-citationArtificial Intelligence in Pancreatic Image Analysis: A Review
Liu, W., Zhang, B., Liu, T., Jiang, J., & Liu, Y. (2024) · Sensors · doi:10.3390/s24144749 · Cited by 35 · see citing works
She, H., Du, Y., Du, Y., Tan, L., Shunxin, Y., Luo, X., Li, Q., Xiang, X., Lu, H., Hu, Y., et al. (2023) · BMC Anesthesiology · doi:10.1186/s12871-023-02317-4 · Cited by 32 · see citing works
Berbis, M. A., Paulano‐Godino, F., Rodríguez‐Comas, J., Baro, E. N., García‐Figueiras, R., Baleato‐González, S., & Luna, A. (2023) · Abdominal Radiology · doi:10.1007/s00261-023-04071-0 · Cited by 31 · see citing works
Early Detection of Pancreatic Cancer: Current Advances and Future Opportunities
Lin, Z., Adeniran, E. A., Cai, Y., Qureshi, T. A., Li, D., Gong, J., Li, J., Pandol, S. J., & Jiang, Y. (2025) · Biomedicines · doi:10.3390/biomedicines13071733 · Cited by 34 · see citing works
Proteomics-Driven Biomarkers in Pancreatic Cancer
Ramalhete, L., Vigia, E., Araújo, R., & Marques, H. P. (2023) · Proteomes · doi:10.3390/proteomes11030024 · Cited by 24 · see citing works
Pancreatic ductal adenocarcinoma staging: a narrative review of radiologic techniques and advances
Chu, L. C. H., & Fishman, E. K. (2023) · International Journal of Surgery · doi:10.1097/js9.0000000000000899 · Cited by 29 · see citing works
Screening for pancreatic cancer has the potential to save lives, but is it practical?
Mazer, B. L., Lee, J. W., Roberts, N. J., Chu, L. C. H., Lennon, A. M., Klein, A. P., Eshleman, J. R., Fishman, E. K., Canto, M. I., Goggins, M., et al. (2023) · Expert Review of Gastroenterology & Hepatology · doi:10.1080/17474124.2023.2217354 · Cited by 23 · see citing works
A Review of the Clinical Applications of Artificial Intelligence in Abdominal Imaging
Mervak, B. M., Fried, J. G., & Wasnik, A. P. (2023) · Diagnostics · doi:10.3390/diagnostics13182889 · Cited by 26 · see citing works
Rai, H. M., Yoo, J., Moqurrab, S. A., & Dashkevych, S. (2023) · Measurement · doi:10.1016/j.measurement.2023.114059 · Cited by 26 · see citing works
The FELIX Project: Deep Networks To Detect Pancreatic NeoplasmsPreprint
Xia, Y., Yu, Q., Chu, L. C. H., Kawamoto, S., Park, S., Liu, F., Chen, J., Zhu, Z., Li, B., Zhou, Z., et al. (2022) · medRxiv · doi:10.1101/2022.09.24.22280071 · Cited by 29 · see citing works
Lee, J. H., Shin, J., Min, J. H., Jeong, W. K., Kim, H., Choi, S.-Y., Lee, J., Hong, S., & Kim, K. (2024) · Cancer Imaging · doi:10.1186/s40644-024-00653-3 · Cited by 28 · see citing works
Janssens, L. P., Takahashi, H., Nagayama, H., Nugen, F., Bamlet, W. R., Oberg, A. L., Fuemmeler, E. G., Goenka, A. H., Erickson, B. J., Takahashi, N., et al. (2023) · Pancreatology · doi:10.1016/j.pan.2023.04.008 · Cited by 25 · see citing works
Self-citationWang, Y., Zhang, L., Jiang, Y., Cheng, X., He, W., Yu, H., Li, X., Yang, J., Yao, G., Lu, Z., et al. (2024) · Quantitative Imaging in Medicine and Surgery · doi:10.21037/qims-24-7 · Cited by 24 · see citing works
The role of biomarkers in the early detection of pancreatic cancer
Goggins, M. (2024) · Familial Cancer · doi:10.1007/s10689-024-00381-4 · Cited by 26 · see citing works
Mukherjee, S., Korfiatis, P. D., Patnam, N. G., Trivedi, K. H., Karbhari, A., Suman, G., Fletcher, J. G., & Goenka, A. H. (2024) · Abdominal Radiology · doi:10.1007/s00261-023-04127-1 · Cited by 20 · see citing works
Self-citationAnghel, C., Grasu, M. C., Anghel, D. A., Rusu-Munteanu, G.-I., Dumitru, R. L., & Lupescu, I. G. (2024) · Diagnostics · doi:10.3390/diagnostics14040438 · Cited by 22 · see citing works
Daamen, L. A., Molenaar, I. Q., & Groot, V. P. (2023) · Journal of Clinical Medicine · doi:10.3390/jcm12237485 · Cited by 20 · see citing works
Schuurmans, M., Alves, N., Vendittelli, P., Huisman, H. J., Hermans, J. J., Litjens, G. J. S., Chang, D. K.-F., Verbeke, C. S., Malats, N., & Löhr, M. J. (2023) · Gastroenterology · doi:10.1053/j.gastro.2023.04.003 · Cited by 20 · see citing works
Challenges of early detection of pancreatic cancer
Shen, M. J., Jamali, A., & Katona, B. W. (2025) · Journal of Clinical Investigation · doi:10.1172/jci191942 · Cited by 22 · see citing works
Cai, W., Wu, X. M., Guo, K., Chen, Y., Shi, Y., & Lin, X. (2024) · Journal of Inflammation Research · doi:10.2147/jir.s484485 · Cited by 20 · see citing works
Tovar, D., Rosenthal, M. H., Maitra, A., & Koay, E. J. (2023) · Artificial Intelligence Surgery · doi:10.20517/ais.2022.38 · Cited by 22 · see citing works
Shen, T., Liu, D., Lin, Z. J., Ren, C., Zhao, W., & Gao, W. (2022) · Journal of Clinical Medicine · doi:10.3390/jcm11206061 · Cited by 18 · see citing works
Lopez-Ramirez, F., Soleimani, S., Azadi, J. R., Sheth, S., Kawamoto, S., Javed, A. A., Tixier, F., Hruban, R. H., Fishman, E. K., & Chu, L. C. H. (2024) · Diagnostic and Interventional Imaging · doi:10.1016/j.diii.2024.08.003 · Cited by 18 · see citing works
PanIN or IPMN? Redefining Lesion Size in 3 Dimensions
Kiemen, A. L., Dequiedt, L., Shen, Y., Zhu, Y., Matos-Romero, V., Forjaz, A., Campbell, K., Dhana, W., Cornish, T. C., Braxton, A. M., et al. (2024) · The American Journal of Surgical Pathology · doi:10.1097/pas.0000000000002245 · Cited by 18 · see citing works
Wang, Y., Li, C., & Wang, Z. (2024) · IEEE Access · doi:10.1109/access.2023.3348810 · Cited by 17 · see citing works
Mukherjee, S., Antony, A., Patnam, N. G., Trivedi, K. H., Karbhari, A., Nagaraj, M., Murlidhar, M., & Goenka, A. H. (2025) · Scientific Reports · doi:10.1038/s41598-025-01802-9 · Cited by 19 · see citing works
Self-citationOyama, H., Hamada, T., Nakai, Y., Tanaka, M. H., Endo, G., Hakuta, R., Ishida, K., Ishigaki, K., Kanai, S., Kurihara, K., et al. (2023) · Journal of Gastroenterology · doi:10.1007/s00535-023-02028-0 · Cited by 16 · see citing works
Mukherjee, S., Korfiatis, P. D., Khasawneh, H., Rajamohan, N., Patra, A., Suman, G., Singh, A., Thakkar, J., Patnam, N. G., Trivedi, K. H., et al. (2023) · Pancreatology · doi:10.1016/j.pan.2023.05.008 · Cited by 15 · see citing works
Self-citationSmith, L. M., Mahoney, D. W., Bamlet, W. R., Yu, F., Liu, S., Goggins, M., Darabi, S., Majumder, S., Wang, Q.-L., Coté, G. A., et al. (2024) · Pancreatology · doi:10.1016/j.pan.2024.10.012 · Cited by 16 · see citing works
Self-citationThanya, T., & Jeslin, T. (2025) · Abdominal Radiology · doi:10.1007/s00261-025-04860-9 · Cited by 13 · see citing works
Huang, C., Hecht, E. M., Soloff, E. V., Tiwari, H. A., Bhosale, P., Dasayam, A., Galgano, S. J., Kambadakone, A., Kulkarni, N. M., Le, O., et al. (2024) · American Journal of Roentgenology · doi:10.2214/ajr.24.31151 · Cited by 13 · see citing works
Self-citationYamada, R., Tsuboi, J., Murashima, Y., Tanaka, T., Nose, K., & Nakagawa, H. (2023) · Biomedicines · doi:10.3390/biomedicines11061687 · Cited by 13 · see citing works
Zhu, J., Xu, B., Fan, T., Ji, S., Gu, K., Ding, J., Lu, H., Ma, J., & Zhou, Y. (2025) · npj Precision Oncology · doi:10.1038/s41698-025-01047-9 · Cited by 13 · see citing works
MicroRNAs as Bile-based biomarkers in pancreaticobiliary cancers (MIRABILE): a cohort study
Liu, D., Puik, J. R., Venø, M. T., et al. (2024) · International Journal of Surgery · doi:10.1097/js9.0000000000001888 · Cited by 12 · see citing works
Imaging of Pancreatic Ductal Adenocarcinoma: An Update on Recent Advances
Barat, M., Marchèse, U., Pellat, A., Dohan, A. J. J., Coriat, R., Hoeffel, C. C., Fishman, E. K., Cassinotto, C., Chu, L. C. H., & Soyer, P. A. (2022) · Canadian Association of Radiologists Journal · doi:10.1177/08465371221124927 · Cited by 12 · see citing works
AI-Driven insights in pancreatic cancer imaging: from pre-diagnostic detection to prognostication
Antony, A., Mukherjee, S., Bi, Y., Collisson, E. A., Nagaraj, M., Murlidhar, M., Wallace, M. B., & Goenka, A. H. (2024) · Abdominal Radiology · doi:10.1007/s00261-024-04775-x · Cited by 11 · see citing works
Self-citationAdvancing pancreatic cancer management: the role of artificial intelligence in diagnosis and therapy
Ozsahin, D. U., Usanase, N., & Özşahin, İ. (2025) · Beni-Suef University Journal of Basic and Applied Sciences · doi:10.1186/s43088-025-00610-4 · Cited by 10 · see citing works
Wright, D., Mukherjee, S., Patra, A., Khasawneh, H., Korfiatis, P. D., Suman, G., Chari, S. T., Kudva, Y. C., Kline, T. L., & Goenka, A. H. (2022) · Abdominal Radiology · doi:10.1007/s00261-022-03668-1 · Cited by 11 · see citing works
Self-citationCT Imaging of the Pancreas: A Review of Current Developments and Applications
Barat, M., Greffier, J., Si-Mohamed, S. A., Dohan, A. J. J., Pellat, A., Frandon, J., Calame, P., & Soyer, P. A. (2025) · Canadian Association of Radiologists Journal · doi:10.1177/08465371251319965 · Cited by 12 · see citing works
Roth, S., Apte, M. V., Balachandran, V. P., Biffi, G., Conroy, T., Cukierman, E., Dotan, E., Esposito, I., Grünwald, B. T., Klauß, M., et al. (2026) · Nature Reviews Disease Primers · doi:10.1038/s41572-026-00699-6 · Cited by 16 · see citing works
Waite, S., Davenport, M. S., Graber, M. L., Banja, J. D., Sheppard, B., & Bruno, M. A. (2024) · American Journal of Roentgenology · doi:10.2214/ajr.24.31686 · Cited by 9 · see citing works
Early detection of pancreatic cancer in the era of precision medicine
Ahmed, T. M., Kawamoto, S., Lopez-Ramirez, F., Yasrab, M., Hruban, R. H., Fishman, E. K., & Chu, L. C. H. (2024) · Abdominal Radiology · doi:10.1007/s00261-024-04358-w · Cited by 11 · see citing works
Renjifo-Correa, M. E., Fanni, S. C., Bustamante–Cristancho, L. A., Cuibari, M. E., Aghakhanyan, G., Faggioni, L., Neri, E., & Cioni, D. (2025) · Cancers · doi:10.3390/cancers17050803 · Cited by 10 · see citing works
Gültekin, M. A., Peker, A. A., Oktay, A. B., et al. (2023) · Journal of Clinical Ultrasound · doi:10.1002/jcu.23558 · Cited by 9 · see citing works
Prävention des Pankreaskarzinoms
Scherübl, H. (2023) · DMW - Deutsche Medizinische Wochenschrift · doi:10.1055/a-1975-2366 · Cited by 8 · see citing works
Advancements in Radiomics-Based AI for Pancreatic Ductal Adenocarcinoma
Lekkas, G., Vrochidou, Ε., & Papakostas, G. A. (2025) · Bioengineering · doi:10.3390/bioengineering12080849 · Cited by 9 · see citing works
Yuan, L., Ji, M., Wang, S., Lu, X., Li, Y., Huang, P., Lu, C. Y., Shen, L., & Xu, J. (2024) · iScience · doi:10.1016/j.isci.2024.111058 · Cited by 8 · see citing works
Mukherjee, S., Antony, A., Patnam, N. G., Trivedi, K. H., Karbhari, A., Bhinder, K. K., Zarrintan, A., Fletcher, J. G., Truty, M. J., Johnson, M. P., et al. (2026) · Gut · doi:10.1136/gutjnl-2025-337266 · Cited by 9 · see citing works
Self-citationTotal Neoadjuvant Therapy in Localized Pancreatic Cancer: Is More Better?
Saúde-Conde, R., Ghali, B. E., Navez, J., Bouchart, C., & Laethem, J. V. (2024) · Cancers · doi:10.3390/cancers16132423 · Cited by 9 · see citing works
Arshad, H., Lopez-Ramirez, F., Tixier, F., Soyer, P. A., Kawamoto, S., Fishman, E. K., & Chu, L. C. H. (2025) · Canadian Association of Radiologists Journal · doi:10.1177/08465371251351810 · Cited by 9 · see citing works
Ye, Y., Zhang, J., Song, P., Qin, P., Hu, Y., An, P., Li, X., Lin, Y., Wang, J., & Feng, G. (2023) · Technology in Cancer Research & Treatment · doi:10.1177/15330338231180792 · Cited by 7 · see citing works
Wang, J., Zhou, Y., Zhou, J., Liu, H., & Li, X. (2024) · European Journal of Radiology · doi:10.1016/j.ejrad.2024.111531 · Cited by 7 · see citing works
Huang, C., Chopra, S., Bolan, C. W., Chandarana, H., Harfouch, N., Hecht, E. M., Lo, G. C., & Megibow, A. J. (2023) · Gastrointestinal Endoscopy Clinics of North America · doi:10.1016/j.giec.2023.03.004 · Cited by 7 · see citing works
Heuriskance: a novel paradigm for systematic earlier detection of sporadic pancreatic cancer
Chari, S. T., Feng, Z., Wu, B. U., Fisher, W., Kambadakone, A. R., Zhao, Y., Maitra, A., Kenner, B. J., & Matrisian, L. M. (2025) · JNCI Journal of the National Cancer Institute · doi:10.1093/jnci/djaf291 · Cited by 6 · see citing works
Self-citationHe, S., Zhao, Y., Shi, L., Yang, X., Wang, X., Luo, Y., Wang, M., Zhang, X., Li, X., Yu, D., et al. (2024) · Scientific Reports · doi:10.1038/s41598-024-70134-x · Cited by 6 · see citing works
Understanding the Conundrum of Pancreatic Cancer in the Omics Sciences Era
Nicoletti, A., Paratore, M., Vitale, F., Negri, M., Quero, G., Esposto, G., Mignini, I., Alfieri, S., Gasbarrini, A., Zocco, M. A., et al. (2024) · International Journal of Molecular Sciences · doi:10.3390/ijms25147623 · Cited by 6 · see citing works
The role of radiomics in hepato-bilio-pancreatic surgery: a literature review
Robertis, R. D., Todesco, M., Autelitano, D., Spoto, F., & D’Onofrio, M. (2023) · Artificial Intelligence Surgery · doi:10.20517/ais.2023.18 · Cited by 6 · see citing works
Applications of artificial intelligence in abdominal imaging
(2025) · doi:10.1007/s00261-025-04990-0 · Cited by 6 · see citing works
Self-citationChen, L., Wang, X., Wang, S., Zhao, X., Yan, Y., Yuan, M., & Sun, S. (2025) · BMC Medical Imaging · doi:10.1186/s12880-025-01722-0 · Cited by 5 · see citing works
Bloomfield, G. C., Shoucair, S., Nigam, A., Park, B. U., Fishbein, T., Radkani, P., & Winslow, E. R. (2024) · Surgery · doi:10.1016/j.surg.2024.06.051 · Cited by 5 · see citing works
Huang, C., Shen, Y., Galgano, S. J., Goenka, A. H., Hecht, E. M., Kambadakone, A., Wang, Z. J., & Chu, L. C. H. (2024) · Abdominal Radiology · doi:10.1007/s00261-024-04644-7 · Cited by 5 · see citing works
Self-citationPancreatic incidentaloma: incidental findings from history towards the era of liquid biopsy
Löhr, M. J., Vujasinović, M., Kartalis, N., & Osten, P. (2024) · eGastroenterology · doi:10.1136/egastro-2024-100082 · Cited by 5 · see citing works
Costa, G. L. L., Petroski, G. T., Machado, L. G., Santos, B. E., Ramos, F. D. O., Feuerschuette, L. M., & Canto, G. D. L. (2024) · Abdominal Radiology · doi:10.1007/s00261-024-04771-1 · Cited by 5 · see citing works
Wang, L., Shi, Y., Qureshi, T. A., Xie, Y., Gaddam, S., Azab, L., Wu, C., He, Y., Deng, Z., Javed, S., et al. (2025) · Cancers · doi:10.3390/cancers17111886 · Cited by 4 · see citing works
Wu, L., Shan, X., Ge, R., et al. (2025) · IEEE International Symposium on Biomedical Imaging (ISBI) · doi:10.1109/isbi60581.2025.10980821 · Cited by 4 · see citing works
Harandi, H., Gouravani, M., Alikarami, S., Farahani, M. S., Ghavam, M., Mohammadi, S., Salehi, M. A., Reynolds, S., Firouzabadi, F. D., & Huda, F. (2025) · Journal of Imaging Informatics in Medicine · doi:10.1007/s10278-025-01607-2 · Cited by 5 · see citing works
Yazarkan, Y., Sönmez, G., Sahin, T. K., Rizzo, A., & Güven, D. C. (2026) · Expert Review of Molecular Diagnostics · doi:10.1080/14737159.2026.2648127 · Cited by 5 · see citing works
The Role of Artificial Intelligence and Big Data for Gastrointestinal Disease
Holt, N., & Byrne, M. F. (2024) · Gastrointestinal Endoscopy Clinics of North America · doi:10.1016/j.giec.2024.09.004 · Cited by 4 · see citing works
Kim, D. H. (2025) · Journal of Personalized Medicine · doi:10.3390/jpm15050175 · Cited by 5 · see citing works
Maurya, R., Chug, I., Vudatha, V., & Palma, A. M. (2024) · Advances in cancer research · doi:10.1016/bs.acr.2024.06.007 · Cited by 3 · see citing works
Intraductal Papillary Mucinous Neoplasm and Pancreatic Cancer: Opportunity Knocks Twice
Hamada, T., Oyama, H., Nakai, Y., Petrov, M. S., & for the TOP-CREATE study group (2025) · The American Journal of Gastroenterology · doi:10.14309/ajg.0000000000003677 · Cited by 3 · see citing works
Gu, Q., Xing, Y., Hu, X., Yang, J., Chen, Y., He, Y., & Liu, P. (2025) · Academic Radiology · doi:10.1016/j.acra.2025.04.025 · Cited by 3 · see citing works
- Authors
- Mukherjee, S., Patra, A., Khasawneh, H., Korfiatis, P., Rajamohan, N., Suman, G., Majumder, S., Panda, A., Johnson, M. P., Larson, N. B., Wright, D. E., Kline, T. L., Fletcher, J. G., Chari, S. T., & Goenka, A. H.
- Affiliations
- Mayo Clinic (US)
- Tata Medical Center (IN)
- Corresponding author
- Goenka, Ajit H.
- ORCID
- Venue
- Gastroenterology
- Publisher
- Elsevier BV
- Published
- 2022-11
- Volume
- 163
- Issue
- 5
- Pages
- 1435-1446.e3
- ISSN
- 0016-5085 (ISSN-L)
- PMC
- PMC12285712
- OpenAlex
- W4283780227
- Semantic Scholar
- 703522bc0ca697394391ed9dd20d76b2e43d6691
- Cited by
- 187
- References
- 36
- Citations by year
- 2026: 35 · 2025: 61 · 2024: 47 · 2023: 35 · 2022: 9
- FWCI
- 21.62
- Type
- journal article
- Language
- en
- Open access
- bronze
- doi.org
- Licence
- Locations
- Gastroenterology (acceptedVersion)
- PubMed (publishedVersion)
- PubMed Central (acceptedVersion)
- Funders
- U.S. Department of Defense
- National Cancer Institute
- Congressionally Directed Medical Research Programs (R01CA256969)
- Mayo Clinic
- Department of Radiology, Weill Cornell Medical College
- National Institutes of Health
- Topics
- Health Sciences › Medicine › Radiology, Nuclear Medicine and Imaging › Radiomics and Machine Learning in Medical Imaging
- Physical Sciences › Computer Science › Artificial Intelligence › AI in cancer detection
- Health Sciences › Medicine › Oncology › Pancreatic and Hepatic Oncology Research
- Keywords
- pancreatic ductal adenocarcinoma, radiomics, early cancer detection, computed tomography, machine learning classification, support vector machine, radiologist comparison, pancreas segmentation, radiomic feature selection, lead time, cancer screening
- MeSH terms
- Machine Learning, Humans, Pancreatic Neoplasms, Retrospective Studies, Tomography, X-Ray Computed, Case-Control Studies, Carcinoma, Pancreatic Ductal
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Claims read under definition version 3 (extract-v3) on 2026-09-28.
Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-09-28.
Cited sources are checked by gpt-6-sol.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
