Related Experiment Video
Updated: Mar 14, 2026
10:38
DIPLOMA Approach for Standardized Pathology Assessment of Distal Pancreatectomy Specimens
Published on: February 1, 2020
8.0K
Evaluating IPMN and pancreatic carcinoma utilizing quantitative histopathology
Evan S Glazer1, Hao Helen Zhang2, Kimberly A Hill3
1University of Tennessee Health Sciences Center, Memphis, Tennessee.
Cancer Medicine
|September 27, 2016
Summary
Quantitative histopathology accurately distinguishes pancreatic lesions. This novel method uses nuclear image signatures to classify chronic pancreatitis (CP), intraductal papillary mucinous neoplasms (IPMN), and pancreatic cancer (PC) with high accuracy.
Area of Science:
- Pathology
- Oncology
- Medical Imaging
Background:
- Intraductal papillary mucinous neoplasms (IPMN) exhibit uncertain biologic behavior, posing diagnostic challenges.
- Accurate classification of pancreatic lesions is crucial for appropriate patient management.
Purpose of the Study:
- To develop objective prediction tools for classifying pancreatic lesions.
- To utilize quantitative histopathological signatures of nuclear images for lesion classification.
Main Methods:
- Retrospective analysis of 44 pancreatic resection patients (12 CP, 16 IPMN, 16 PC).
- Application of regularized multinomial regression for automated, blinded classification of specimens.
- Quantitative analysis of nuclear images, with 180 ± 22 nuclei imaged per lesion.
Main Results:
- Overall classification accuracy reached 89.6% using six unique nuclear features.
- No chronic pancreatitis (CP) cases were misclassified; 1 IPMN and 4 PC cases were misclassified.
- A quantitative measure of classification certainty was developed, with lower certainty correlating with misclassification (P=0.0005).
Conclusions:
- Quantitative histopathology offers a robust method for distinguishing between CP, IPMN, and PC.
- The developed method provides a quantitative measure of diagnostic uncertainty.
- This approach may aid in cases with diagnostic ambiguity.
