A deep learning model to triage and predict adenocarcinoma on pancreas cytology whole slide imaging
Abstract Pancreatic fine-needle aspirations are the gold-standard diagnostic procedure for the evaluation of pancreatic ductal adenocarcinoma. A suspicion for malignancy can escalate towards chemotherapy followed by a major surgery and therefore is a high-stakes task for the pathologist. In this pap...
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2023-10-01
|
| Series: | Scientific Reports |
| Online Access: | https://doi.org/10.1038/s41598-023-42045-w |
| _version_ | 1797453108452786176 |
|---|---|
| author | Andrew Sohn Daniel Miller Efrain Ribeiro Nakul Shankar Syed Ali Ralph Hruban Alexander Baras |
| author_facet | Andrew Sohn Daniel Miller Efrain Ribeiro Nakul Shankar Syed Ali Ralph Hruban Alexander Baras |
| author_sort | Andrew Sohn |
| collection | DOAJ |
| description | Abstract Pancreatic fine-needle aspirations are the gold-standard diagnostic procedure for the evaluation of pancreatic ductal adenocarcinoma. A suspicion for malignancy can escalate towards chemotherapy followed by a major surgery and therefore is a high-stakes task for the pathologist. In this paper, we propose a deep learning framework, MIPCL, that can serve as a helpful screening tool, predicting the presence or absence of cancer. We also reproduce two deep learning models that have found success in surgical pathology for our cytopathology study. Our MIPCL significantly improves over both models across all evaluated metrics (F1-Score: 87.97% vs 88.70% vs 91.07%; AUROC: 0.9159 vs. 0.9051 vs 0.9435). Additionally, our model is able to recover the most contributing regions on the slide for the final prediction. We also present a dataset curation strategy that increases the number of training examples from an existing dataset, thereby reducing the resource burden tied to collecting and scanning additional cases. |
| first_indexed | 2024-03-09T15:18:09Z |
| format | Article |
| id | doaj.art-0399f8578b4b4c50abf7625e2d151481 |
| institution | Directory Open Access Journal |
| issn | 2045-2322 |
| language | English |
| last_indexed | 2024-03-09T15:18:09Z |
| publishDate | 2023-10-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Scientific Reports |
| spelling | doaj.art-0399f8578b4b4c50abf7625e2d1514812023-11-26T12:58:07ZengNature PortfolioScientific Reports2045-23222023-10-0113111310.1038/s41598-023-42045-wA deep learning model to triage and predict adenocarcinoma on pancreas cytology whole slide imagingAndrew Sohn0Daniel Miller1Efrain Ribeiro2Nakul Shankar3Syed Ali4Ralph Hruban5Alexander Baras6Department of Pathology, Johns Hopkins University School of MedicineDepartment of Pathology, Saint Louis University School of MedicineDepartment of Pathology, Johns Hopkins University School of MedicineDepartment of Pathology, University of ColoradoDepartment of Pathology, Johns Hopkins University School of MedicineDepartment of Pathology, Johns Hopkins University School of MedicineDepartment of Pathology, Johns Hopkins University School of MedicineAbstract Pancreatic fine-needle aspirations are the gold-standard diagnostic procedure for the evaluation of pancreatic ductal adenocarcinoma. A suspicion for malignancy can escalate towards chemotherapy followed by a major surgery and therefore is a high-stakes task for the pathologist. In this paper, we propose a deep learning framework, MIPCL, that can serve as a helpful screening tool, predicting the presence or absence of cancer. We also reproduce two deep learning models that have found success in surgical pathology for our cytopathology study. Our MIPCL significantly improves over both models across all evaluated metrics (F1-Score: 87.97% vs 88.70% vs 91.07%; AUROC: 0.9159 vs. 0.9051 vs 0.9435). Additionally, our model is able to recover the most contributing regions on the slide for the final prediction. We also present a dataset curation strategy that increases the number of training examples from an existing dataset, thereby reducing the resource burden tied to collecting and scanning additional cases.https://doi.org/10.1038/s41598-023-42045-w |
| spellingShingle | Andrew Sohn Daniel Miller Efrain Ribeiro Nakul Shankar Syed Ali Ralph Hruban Alexander Baras A deep learning model to triage and predict adenocarcinoma on pancreas cytology whole slide imaging Scientific Reports |
| title | A deep learning model to triage and predict adenocarcinoma on pancreas cytology whole slide imaging |
| title_full | A deep learning model to triage and predict adenocarcinoma on pancreas cytology whole slide imaging |
| title_fullStr | A deep learning model to triage and predict adenocarcinoma on pancreas cytology whole slide imaging |
| title_full_unstemmed | A deep learning model to triage and predict adenocarcinoma on pancreas cytology whole slide imaging |
| title_short | A deep learning model to triage and predict adenocarcinoma on pancreas cytology whole slide imaging |
| title_sort | deep learning model to triage and predict adenocarcinoma on pancreas cytology whole slide imaging |
| url | https://doi.org/10.1038/s41598-023-42045-w |
| work_keys_str_mv | AT andrewsohn adeeplearningmodeltotriageandpredictadenocarcinomaonpancreascytologywholeslideimaging AT danielmiller adeeplearningmodeltotriageandpredictadenocarcinomaonpancreascytologywholeslideimaging AT efrainribeiro adeeplearningmodeltotriageandpredictadenocarcinomaonpancreascytologywholeslideimaging AT nakulshankar adeeplearningmodeltotriageandpredictadenocarcinomaonpancreascytologywholeslideimaging AT syedali adeeplearningmodeltotriageandpredictadenocarcinomaonpancreascytologywholeslideimaging AT ralphhruban adeeplearningmodeltotriageandpredictadenocarcinomaonpancreascytologywholeslideimaging AT alexanderbaras adeeplearningmodeltotriageandpredictadenocarcinomaonpancreascytologywholeslideimaging AT andrewsohn deeplearningmodeltotriageandpredictadenocarcinomaonpancreascytologywholeslideimaging AT danielmiller deeplearningmodeltotriageandpredictadenocarcinomaonpancreascytologywholeslideimaging AT efrainribeiro deeplearningmodeltotriageandpredictadenocarcinomaonpancreascytologywholeslideimaging AT nakulshankar deeplearningmodeltotriageandpredictadenocarcinomaonpancreascytologywholeslideimaging AT syedali deeplearningmodeltotriageandpredictadenocarcinomaonpancreascytologywholeslideimaging AT ralphhruban deeplearningmodeltotriageandpredictadenocarcinomaonpancreascytologywholeslideimaging AT alexanderbaras deeplearningmodeltotriageandpredictadenocarcinomaonpancreascytologywholeslideimaging |