1
|
Dominguez-Morales JP, Duran-Lopez L, Marini N, Vicente-Diaz S, Linares-Barranco A, Atzori M, Müller H. A systematic comparison of deep learning methods for Gleason grading and scoring. Med Image Anal 2024; 95:103191. [PMID: 38728903 DOI: 10.1016/j.media.2024.103191] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 05/10/2022] [Revised: 01/16/2024] [Accepted: 05/02/2024] [Indexed: 05/12/2024]
Abstract
Prostate cancer is the second most frequent cancer in men worldwide after lung cancer. Its diagnosis is based on the identification of the Gleason score that evaluates the abnormality of cells in glands through the analysis of the different Gleason patterns within tissue samples. The recent advancements in computational pathology, a domain aiming at developing algorithms to automatically analyze digitized histopathology images, lead to a large variety and availability of datasets and algorithms for Gleason grading and scoring. However, there is no clear consensus on which methods are best suited for each problem in relation to the characteristics of data and labels. This paper provides a systematic comparison on nine datasets with state-of-the-art training approaches for deep neural networks (including fully-supervised learning, weakly-supervised learning, semi-supervised learning, Additive-MIL, Attention-Based MIL, Dual-Stream MIL, TransMIL and CLAM) applied to Gleason grading and scoring tasks. The nine datasets are collected from pathology institutes and openly accessible repositories. The results show that the best methods for Gleason grading and Gleason scoring tasks are fully supervised learning and CLAM, respectively, guiding researchers to the best practice to adopt depending on the task to solve and the labels that are available.
Collapse
Affiliation(s)
- Juan P Dominguez-Morales
- Robotics and Technology of Computers Lab., ETSII-EPS, Universidad de Sevilla, Sevilla 41012, Spain; SCORE Lab, I3US. Universidad de Sevilla, Spain.
| | - Lourdes Duran-Lopez
- Robotics and Technology of Computers Lab., ETSII-EPS, Universidad de Sevilla, Sevilla 41012, Spain; SCORE Lab, I3US. Universidad de Sevilla, Spain
| | - Niccolò Marini
- Information Systems Institute, University of Applied Sciences Western Switzerland (HES-SO Valais), Technopôle 3, Sierre 3960, Switzerland; Centre Universitaire d'Informatique, University of Geneva, Carouge 1227, Switzerland
| | - Saturnino Vicente-Diaz
- Robotics and Technology of Computers Lab., ETSII-EPS, Universidad de Sevilla, Sevilla 41012, Spain; SCORE Lab, I3US. Universidad de Sevilla, Spain
| | - Alejandro Linares-Barranco
- Robotics and Technology of Computers Lab., ETSII-EPS, Universidad de Sevilla, Sevilla 41012, Spain; SCORE Lab, I3US. Universidad de Sevilla, Spain
| | - Manfredo Atzori
- Information Systems Institute, University of Applied Sciences Western Switzerland (HES-SO Valais), Technopôle 3, Sierre 3960, Switzerland; Department of Neuroscience, University of Padua, Via Giustiniani 2, Padua, 35128, Italy
| | - Henning Müller
- Information Systems Institute, University of Applied Sciences Western Switzerland (HES-SO Valais), Technopôle 3, Sierre 3960, Switzerland; Medical faculty, University of Geneva, Geneva 1211, Switzerland
| |
Collapse
|
2
|
Frewing A, Gibson AB, Robertson R, Urie PM, Corte DD. Don't Fear the Artificial Intelligence: A Systematic Review of Machine Learning for Prostate Cancer Detection in Pathology. Arch Pathol Lab Med 2024; 148:603-612. [PMID: 37594900 DOI: 10.5858/arpa.2022-0460-ra] [Citation(s) in RCA: 1] [Impact Index Per Article: 1.0] [Reference Citation Analysis] [Abstract] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Accepted: 05/04/2023] [Indexed: 08/20/2023]
Abstract
CONTEXT Automated prostate cancer detection using machine learning technology has led to speculation that pathologists will soon be replaced by algorithms. This review covers the development of machine learning algorithms and their reported effectiveness specific to prostate cancer detection and Gleason grading. OBJECTIVE To examine current algorithms regarding their accuracy and classification abilities. We provide a general explanation of the technology and how it is being used in clinical practice. The challenges to the application of machine learning algorithms in clinical practice are also discussed. DATA SOURCES The literature for this review was identified and collected using a systematic search. Criteria were established prior to the sorting process to effectively direct the selection of studies. A 4-point system was implemented to rank the papers according to their relevancy. For papers accepted as relevant to our metrics, all cited and citing studies were also reviewed. Studies were then categorized based on whether they implemented binary or multi-class classification methods. Data were extracted from papers that contained accuracy, area under the curve (AUC), or κ values in the context of prostate cancer detection. The results were visually summarized to present accuracy trends between classification abilities. CONCLUSIONS It is more difficult to achieve high accuracy metrics for multiclassification tasks than for binary tasks. The clinical implementation of an algorithm that can assign a Gleason grade to clinical whole slide images (WSIs) remains elusive. Machine learning technology is currently not able to replace pathologists but can serve as an important safeguard against misdiagnosis.
Collapse
Affiliation(s)
- Aaryn Frewing
- From the Department of Physics and Astronomy, Brigham Young University, Provo, Utah
| | - Alexander B Gibson
- From the Department of Physics and Astronomy, Brigham Young University, Provo, Utah
| | - Richard Robertson
- From the Department of Physics and Astronomy, Brigham Young University, Provo, Utah
| | - Paul M Urie
- From the Department of Physics and Astronomy, Brigham Young University, Provo, Utah
| | - Dennis Della Corte
- From the Department of Physics and Astronomy, Brigham Young University, Provo, Utah
| |
Collapse
|
3
|
Zhu L, Pan J, Mou W, Deng L, Zhu Y, Wang Y, Pareek G, Hyams E, Carneiro BA, Hadfield MJ, El-Deiry WS, Yang T, Tan T, Tong T, Ta N, Zhu Y, Gao Y, Lai Y, Cheng L, Chen R, Xue W. Harnessing artificial intelligence for prostate cancer management. Cell Rep Med 2024; 5:101506. [PMID: 38593808 PMCID: PMC11031422 DOI: 10.1016/j.xcrm.2024.101506] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 08/30/2023] [Revised: 01/05/2024] [Accepted: 03/19/2024] [Indexed: 04/11/2024]
Abstract
Prostate cancer (PCa) is a common malignancy in males. The pathology review of PCa is crucial for clinical decision-making, but traditional pathology review is labor intensive and subjective to some extent. Digital pathology and whole-slide imaging enable the application of artificial intelligence (AI) in pathology. This review highlights the success of AI in detecting and grading PCa, predicting patient outcomes, and identifying molecular subtypes. We propose that AI-based methods could collaborate with pathologists to reduce workload and assist clinicians in formulating treatment recommendations. We also introduce the general process and challenges in developing AI pathology models for PCa. Importantly, we summarize publicly available datasets and open-source codes to facilitate the utilization of existing data and the comparison of the performance of different models to improve future studies.
Collapse
Affiliation(s)
- Lingxuan Zhu
- Department of Urology, Renji Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai 200127, China; Department of Etiology and Carcinogenesis, National Cancer Center/National Clinical Research Center for Cancer/Cancer Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, China; Changping Laboratory, Beijing, China
| | - Jiahua Pan
- Department of Urology, Renji Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai 200127, China
| | - Weiming Mou
- Department of Urology, Shanghai General Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai, China
| | - Longxin Deng
- Department of Urology, Shanghai Changhai Hospital, Second Military Medical University, Shanghai 200433, China
| | - Yinjie Zhu
- Department of Urology, Renji Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai 200127, China
| | - Yanqing Wang
- Department of Urology, Renji Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai 200127, China
| | - Gyan Pareek
- Department of Surgery (Urology), Brown University Warren Alpert Medical School, Providence, RI, USA; Minimally Invasive Urology Institute, Providence, RI, USA
| | - Elias Hyams
- Department of Surgery (Urology), Brown University Warren Alpert Medical School, Providence, RI, USA; Minimally Invasive Urology Institute, Providence, RI, USA
| | - Benedito A Carneiro
- The Legorreta Cancer Center at Brown University, Lifespan Cancer Institute, Providence, RI, USA
| | - Matthew J Hadfield
- The Legorreta Cancer Center at Brown University, Lifespan Cancer Institute, Providence, RI, USA
| | - Wafik S El-Deiry
- The Legorreta Cancer Center at Brown University, Laboratory of Translational Oncology and Experimental Cancer Therapeutics, Department of Pathology & Laboratory Medicine, The Warren Alpert Medical School of Brown University, The Joint Program in Cancer Biology, Brown University and Lifespan Health System, Division of Hematology/Oncology, The Warren Alpert Medical School of Brown University, Providence, RI, USA
| | - Tao Yang
- Department of Medical Oncology, National Cancer Center/National Clinical Research Center for Cancer/Cancer Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, China
| | - Tao Tan
- Faculty of Applied Sciences, Macao Polytechnic University, Address: R. de Luís Gonzaga Gomes, Macao, China
| | - Tong Tong
- College of Physics and Information Engineering, Fuzhou University, Fujian 350108, China
| | - Na Ta
- Department of Pathology, Shanghai Changhai Hospital, Second Military Medical University, Shanghai 200433, China
| | - Yan Zhu
- Department of Pathology, Shanghai Changhai Hospital, Second Military Medical University, Shanghai 200433, China
| | - Yisha Gao
- Department of Pathology, Shanghai Changhai Hospital, Second Military Medical University, Shanghai 200433, China
| | - Yancheng Lai
- Department of Urology, Renji Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai 200127, China; The First School of Clinical Medicine, Southern Medical University, Guangzhou, China
| | - Liang Cheng
- Department of Surgery (Urology), Brown University Warren Alpert Medical School, Providence, RI, USA; Department of Pathology and Laboratory Medicine, Department of Surgery (Urology), Brown University Warren Alpert Medical School, Lifespan Health, and the Legorreta Cancer Center at Brown University, Providence, RI, USA.
| | - Rui Chen
- Department of Urology, Renji Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai 200127, China.
| | - Wei Xue
- Department of Urology, Renji Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai 200127, China.
| |
Collapse
|
4
|
Despotovic V, Kim SY, Hau AC, Kakoichankava A, Klamminger GG, Borgmann FBK, Frauenknecht KB, Mittelbronn M, Nazarov PV. Glioma subtype classification from histopathological images using in-domain and out-of-domain transfer learning: An experimental study. Heliyon 2024; 10:e27515. [PMID: 38562501 PMCID: PMC10982966 DOI: 10.1016/j.heliyon.2024.e27515] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 10/23/2023] [Revised: 02/29/2024] [Accepted: 03/01/2024] [Indexed: 04/04/2024] Open
Abstract
We provide in this paper a comprehensive comparison of various transfer learning strategies and deep learning architectures for computer-aided classification of adult-type diffuse gliomas. We evaluate the generalizability of out-of-domain ImageNet representations for a target domain of histopathological images, and study the impact of in-domain adaptation using self-supervised and multi-task learning approaches for pretraining the models using the medium-to-large scale datasets of histopathological images. A semi-supervised learning approach is furthermore proposed, where the fine-tuned models are utilized to predict the labels of unannotated regions of the whole slide images (WSI). The models are subsequently retrained using the ground-truth labels and weak labels determined in the previous step, providing superior performance in comparison to standard in-domain transfer learning with balanced accuracy of 96.91% and F1-score 97.07%, and minimizing the pathologist's efforts for annotation. Finally, we provide a visualization tool working at WSI level which generates heatmaps that highlight tumor areas; thus, providing insights to pathologists concerning the most informative parts of the WSI.
Collapse
Affiliation(s)
- Vladimir Despotovic
- Bioinformatics Platform, Department of Medical Informatics, Luxembourg Institute of Health, Strassen, Luxembourg
| | - Sang-Yoon Kim
- Bioinformatics Platform, Department of Medical Informatics, Luxembourg Institute of Health, Strassen, Luxembourg
| | - Ann-Christin Hau
- Dr. Senckenberg Institute of Neurooncology, University Hospital Frankfurt, Frankfurt am Main, Germany
- Edinger Institute, Institute of Neurology, Goethe University, Frankfurt am Main, Germany
- Frankfurt Cancer Institute, Goethe University, Frankfurt am Main, Germany
- University Cancer Center Frankfurt, Frankfurt am Main, Germany
- University Hospital, Goethe University, Frankfurt am Main, Germany
- Laboratoire national de santé, National Center of Pathology, Dudelange, Luxembourg
| | - Aliaksandra Kakoichankava
- Multi-Omics Data Science group, Department of Cancer Research, Luxembourg Institute of Health, Strassen, Luxembourg
| | - Gilbert Georg Klamminger
- Luxembourg Centre of Neuropathology, Dudelange, Luxembourg
- Klinik für Frauenheilkunde, Geburtshilfe und Reproduktionsmedizin, Saarland University, Homburg, Germany
| | - Felix Bruno Kleine Borgmann
- Luxembourg Centre of Neuropathology, Dudelange, Luxembourg
- Department of Cancer Research, Luxembourg Institute of Health, Strassen, Luxembourg
- Haupitaux Robert Schumann, Kirchberg, Luxembourg
| | - Katrin B.M. Frauenknecht
- Laboratoire national de santé, National Center of Pathology, Dudelange, Luxembourg
- Luxembourg Centre of Neuropathology, Dudelange, Luxembourg
| | - Michel Mittelbronn
- Laboratoire national de santé, National Center of Pathology, Dudelange, Luxembourg
- Luxembourg Centre of Neuropathology, Dudelange, Luxembourg
- Department of Cancer Research, Luxembourg Institute of Health, Strassen, Luxembourg
- Luxembourg Centre for Systems Biomedicine, University of Luxembourg, Belval, Luxembourg
- Department of Life Sciences and Medicine, University of Luxembourg, Esch-sur-Alzette, Luxembourg
- Faculty of Science, Technology and Medicine, University of Luxembourg, Esch-sur-Alzette, Luxembourg
| | - Petr V. Nazarov
- Bioinformatics Platform, Department of Medical Informatics, Luxembourg Institute of Health, Strassen, Luxembourg
- Multi-Omics Data Science group, Department of Cancer Research, Luxembourg Institute of Health, Strassen, Luxembourg
| |
Collapse
|
5
|
Rabilloud N, Allaume P, Acosta O, De Crevoisier R, Bourgade R, Loussouarn D, Rioux-Leclercq N, Khene ZE, Mathieu R, Bensalah K, Pecot T, Kammerer-Jacquet SF. Deep Learning Methodologies Applied to Digital Pathology in Prostate Cancer: A Systematic Review. Diagnostics (Basel) 2023; 13:2676. [PMID: 37627935 PMCID: PMC10453406 DOI: 10.3390/diagnostics13162676] [Citation(s) in RCA: 4] [Impact Index Per Article: 4.0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 07/25/2023] [Revised: 08/09/2023] [Accepted: 08/11/2023] [Indexed: 08/27/2023] Open
Abstract
Deep learning (DL), often called artificial intelligence (AI), has been increasingly used in Pathology thanks to the use of scanners to digitize slides which allow us to visualize them on monitors and process them with AI algorithms. Many articles have focused on DL applied to prostate cancer (PCa). This systematic review explains the DL applications and their performances for PCa in digital pathology. Article research was performed using PubMed and Embase to collect relevant articles. A Risk of Bias (RoB) was assessed with an adaptation of the QUADAS-2 tool. Out of the 77 included studies, eight focused on pre-processing tasks such as quality assessment or staining normalization. Most articles (n = 53) focused on diagnosis tasks like cancer detection or Gleason grading. Fifteen articles focused on prediction tasks, such as recurrence prediction or genomic correlations. Best performances were reached for cancer detection with an Area Under the Curve (AUC) up to 0.99 with algorithms already available for routine diagnosis. A few biases outlined by the RoB analysis are often found in these articles, such as the lack of external validation. This review was registered on PROSPERO under CRD42023418661.
Collapse
Affiliation(s)
- Noémie Rabilloud
- Impact TEAM, Laboratoire Traitement du Signal et de l’Image (LTSI) INSERM, Rennes University, 35033 Rennes, France (S.-F.K.-J.)
| | - Pierre Allaume
- Department of Pathology, Rennes University Hospital, 2 rue Henri Le Guilloux, CEDEX 09, 35033 Rennes, France; (P.A.)
| | - Oscar Acosta
- Impact TEAM, Laboratoire Traitement du Signal et de l’Image (LTSI) INSERM, Rennes University, 35033 Rennes, France (S.-F.K.-J.)
| | - Renaud De Crevoisier
- Impact TEAM, Laboratoire Traitement du Signal et de l’Image (LTSI) INSERM, Rennes University, 35033 Rennes, France (S.-F.K.-J.)
- Department of Radiotherapy, Centre Eugène Marquis, 35033 Rennes, France
| | - Raphael Bourgade
- Department of Pathology, Nantes University Hospital, 44000 Nantes, France
| | | | - Nathalie Rioux-Leclercq
- Department of Pathology, Rennes University Hospital, 2 rue Henri Le Guilloux, CEDEX 09, 35033 Rennes, France; (P.A.)
| | - Zine-eddine Khene
- Impact TEAM, Laboratoire Traitement du Signal et de l’Image (LTSI) INSERM, Rennes University, 35033 Rennes, France (S.-F.K.-J.)
- Department of Urology, Rennes University Hospital, 2 rue Henri Le Guilloux, CEDEX 09, 35033 Rennes, France
| | - Romain Mathieu
- Department of Urology, Rennes University Hospital, 2 rue Henri Le Guilloux, CEDEX 09, 35033 Rennes, France
| | - Karim Bensalah
- Department of Urology, Rennes University Hospital, 2 rue Henri Le Guilloux, CEDEX 09, 35033 Rennes, France
| | - Thierry Pecot
- Facility for Artificial Intelligence and Image Analysis (FAIIA), Biosit UAR 3480 CNRS-US18 INSERM, Rennes University, 2 Avenue du Professeur Léon Bernard, 35042 Rennes, France
| | - Solene-Florence Kammerer-Jacquet
- Impact TEAM, Laboratoire Traitement du Signal et de l’Image (LTSI) INSERM, Rennes University, 35033 Rennes, France (S.-F.K.-J.)
- Department of Pathology, Rennes University Hospital, 2 rue Henri Le Guilloux, CEDEX 09, 35033 Rennes, France; (P.A.)
| |
Collapse
|
6
|
Bagcilar O, Alis D, Alis C, Seker ME, Yergin M, Ustundag A, Hikmet E, Tezcan A, Polat G, Akkus AT, Alper F, Velioglu M, Yildiz O, Selcuk HH, Oksuz I, Kizilkilic O, Karaarslan E. Automated LVO detection and collateral scoring on CTA using a 3D self-configuring object detection network: a multi-center study. Sci Rep 2023; 13:8834. [PMID: 37258516 DOI: 10.1038/s41598-023-33723-w] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 12/30/2022] [Accepted: 04/18/2023] [Indexed: 06/02/2023] Open
Abstract
The use of deep learning (DL) techniques for automated diagnosis of large vessel occlusion (LVO) and collateral scoring on computed tomography angiography (CTA) is gaining attention. In this study, a state-of-the-art self-configuring object detection network called nnDetection was used to detect LVO and assess collateralization on CTA scans using a multi-task 3D object detection approach. The model was trained on single-phase CTA scans of 2425 patients at five centers, and its performance was evaluated on an external test set of 345 patients from another center. Ground-truth labels for the presence of LVO and collateral scores were provided by three radiologists. The nnDetection model achieved a diagnostic accuracy of 98.26% (95% CI 96.25-99.36%) in identifying LVO, correctly classifying 339 out of 345 CTA scans in the external test set. The DL-based collateral scores had a kappa of 0.80, indicating good agreement with the consensus of the radiologists. These results demonstrate that the self-configuring 3D nnDetection model can accurately detect LVO on single-phase CTA scans and provide semi-quantitative collateral scores, offering a comprehensive approach for automated stroke diagnostics in patients with LVO.
Collapse
Affiliation(s)
- Omer Bagcilar
- Radiology Department, Sisli Hamidiye Etfal Research and Training Hospital, Istanbul, Turkey
| | - Deniz Alis
- Radiology Department, School of Medicine, Acibadem Mehmet Ali Aydinlar University, Istanbul, Turkey.
- Artificial Intelligence, and Information Technologies, Hevi AI Health, Istanbul, Turkey.
| | - Ceren Alis
- Neurology Department, Istanbul Istinye State Hospital, Istanbul, Turkey
| | - Mustafa Ege Seker
- School of Medicine, Acibadem Mehmet Ali Aydinlar University, Istanbul, Turkey
| | - Mert Yergin
- Artificial Intelligence, and Information Technologies, Hevi AI Health, Istanbul, Turkey
| | - Ahmet Ustundag
- Radiology Department, Cerrahpaşa Medical Faculty, Istanbul University-Cerrahpasa, Istanbul, Turkey
| | - Emil Hikmet
- Radiology Department, Cerrahpaşa Medical Faculty, Istanbul University-Cerrahpasa, Istanbul, Turkey
| | - Alperen Tezcan
- Radiology Department, School of Medicine, Erzurum Ataturk University, Istanbul, Turkey
| | - Gokhan Polat
- Radiology Department, School of Medicine, Erzurum Ataturk University, Istanbul, Turkey
| | - Ahmet Tugrul Akkus
- Radiology Department, School of Medicine, Erzurum Ataturk University, Istanbul, Turkey
| | - Fatih Alper
- Radiology Department, School of Medicine, Erzurum Ataturk University, Istanbul, Turkey
| | - Murat Velioglu
- Radiology Department, Istanbul Fatih Sultan Mehmet Training and Research Hospital, Istanbul, Turkey
| | - Omer Yildiz
- Radiology Department, Istanbul Fatih Sultan Mehmet Training and Research Hospital, Istanbul, Turkey
| | - Hakan Hatem Selcuk
- Radiology Department, Istanbul Bakırköy Sadi Konuk Training and Research Hospital, Istanbul, Turkey
| | - Ilkay Oksuz
- Computer Engineering Department, Istanbul Technical University, Istanbul, Turkey
| | - Osman Kizilkilic
- Radiology Department, Cerrahpaşa Medical Faculty, Istanbul University-Cerrahpasa, Istanbul, Turkey
| | - Ercan Karaarslan
- Radiology Department, School of Medicine, Acibadem Mehmet Ali Aydinlar University, Istanbul, Turkey
| |
Collapse
|
7
|
Tsuneki M, Abe M, Ichihara S, Kanavati F. Inference of core needle biopsy whole slide images requiring definitive therapy for prostate cancer. BMC Cancer 2023; 23:11. [PMID: 36600203 DOI: 10.1186/s12885-022-10488-5] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 09/16/2022] [Accepted: 12/26/2022] [Indexed: 01/06/2023] Open
Abstract
BACKGROUND Prostate cancer is often a slowly progressive indolent disease. Unnecessary treatments from overdiagnosis are a significant concern, particularly low-grade disease. Active surveillance has being considered as a risk management strategy to avoid potential side effects by unnecessary radical treatment. In 2016, American Society of Clinical Oncology (ASCO) endorsed the Cancer Care Ontario (CCO) Clinical Practice Guideline on active surveillance for the management of localized prostate cancer. METHODS Based on this guideline, we developed a deep learning model to classify prostate adenocarcinoma into indolent (applicable for active surveillance) and aggressive (necessary for definitive therapy) on core needle biopsy whole slide images (WSIs). In this study, we trained deep learning models using a combination of transfer, weakly supervised, and fully supervised learning approaches using a dataset of core needle biopsy WSIs (n=1300). In addition, we performed an inter-rater reliability evaluation on the WSI classification. RESULTS We evaluated the models on a test set (n=645), achieving ROC-AUCs of 0.846 for indolent and 0.980 for aggressive. The inter-rater reliability evaluation showed s-scores in the range of 0.10 to 0.95, with the lowest being on the WSIs with both indolent and aggressive classification by the model, and the highest on benign WSIs. CONCLUSION The results demonstrate the promising potential of deployment in a practical prostate adenocarcinoma histopathological diagnostic workflow system.
Collapse
Affiliation(s)
- Masayuki Tsuneki
- Medmain Research, Medmain Inc., 2-4-5-104, Akasaka, Chuo-ku, Fukuoka, 810-0042, Japan.
| | - Makoto Abe
- Department of Pathology, Tochigi Cancer Center, 4-9-13 Yohnan, Utsunomiya, 320-0834, Japan
| | - Shin Ichihara
- Department of Surgical Pathology, Sapporo Kosei General Hospital, 8-5 Kita-3-jo Higashi, Chuo-ku, Sapporo, 060-0033, Japan
| | - Fahdi Kanavati
- Medmain Research, Medmain Inc., 2-4-5-104, Akasaka, Chuo-ku, Fukuoka, 810-0042, Japan
| |
Collapse
|
8
|
Ramamurthy K, Varikuti AR, Gupta B, Aswani N. A deep learning network for Gleason grading of prostate biopsies using EfficientNet. BIOMED ENG-BIOMED TE 2022; 68:187-198. [PMID: 36332194 DOI: 10.1515/bmt-2022-0201] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 05/18/2022] [Accepted: 10/23/2022] [Indexed: 11/06/2022]
Abstract
Abstract
Objectives
The most crucial part in the diagnosis of cancer is severity grading. Gleason’s score is a widely used grading system for prostate cancer. Manual examination of the microscopic images and grading them is tiresome and consumes a lot of time. Hence to automate the Gleason grading process, a novel deep learning network is proposed in this work.
Methods
In this work, a deep learning network for Gleason grading of prostate cancer is proposed based on EfficientNet architecture. It applies a compound scaling method to balance the dimensions of the underlying network. Also, an additional attention branch is added to EfficientNet-B7 for precise feature weighting.
Result
To the best of our knowledge, this is the first work that integrates an additional attention branch with EfficientNet architecture for Gleason grading. The proposed models were trained using H&E-stained samples from prostate cancer Tissue Microarrays (TMAs) in the Harvard Dataverse dataset.
Conclusions
The proposed network was able to outperform the existing methods and it achieved an Kappa score of 0.5775.
Collapse
Affiliation(s)
- Karthik Ramamurthy
- Centre for Cyber Physical Systems, School of Electronics Engineering, Vellore Institute of Technology , Chennai , India
| | - Abinash Reddy Varikuti
- School of Computer Science Engineering, Vellore Institute of Technology , Chennai , India
| | - Bhavya Gupta
- School of Computer Science Engineering, Vellore Institute of Technology , Chennai , India
| | - Nehal Aswani
- School of Electronics Engineering, Vellore Institute of Technology , Chennai , India
| |
Collapse
|
9
|
A Deep Learning Model for Prostate Adenocarcinoma Classification in Needle Biopsy Whole-Slide Images Using Transfer Learning. Diagnostics (Basel) 2022; 12:diagnostics12030768. [PMID: 35328321 PMCID: PMC8947489 DOI: 10.3390/diagnostics12030768] [Citation(s) in RCA: 7] [Impact Index Per Article: 3.5] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 02/18/2022] [Revised: 03/08/2022] [Accepted: 03/18/2022] [Indexed: 02/04/2023] Open
Abstract
The histopathological diagnosis of prostate adenocarcinoma in needle biopsy specimens is of pivotal importance for determining optimum prostate cancer treatment. Since diagnosing a large number of cases containing 12 core biopsy specimens by pathologists using a microscope is time-consuming manual system and limited in terms of human resources, it is necessary to develop new techniques that can rapidly and accurately screen large numbers of histopathological prostate needle biopsy specimens. Computational pathology applications that can assist pathologists in detecting and classifying prostate adenocarcinoma from whole-slide images (WSIs) would be of great benefit for routine pathological practice. In this paper, we trained deep learning models capable of classifying needle biopsy WSIs into adenocarcinoma and benign (non-neoplastic) lesions. We evaluated the models on needle biopsy, transurethral resection of the prostate (TUR-P), and The Cancer Genome Atlas (TCGA) public dataset test sets, achieving an ROC-AUC up to 0.978 in needle biopsy test sets and up to 0.9873 in TCGA test sets for adenocarcinoma.
Collapse
|
10
|
Semi-supervised training of deep convolutional neural networks with heterogeneous data and few local annotations: An experiment on prostate histopathology image classification. Med Image Anal 2021; 73:102165. [PMID: 34303169 DOI: 10.1016/j.media.2021.102165] [Citation(s) in RCA: 26] [Impact Index Per Article: 8.7] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 08/07/2020] [Revised: 06/29/2021] [Accepted: 07/06/2021] [Indexed: 01/18/2023]
Abstract
Convolutional neural networks (CNNs) are state-of-the-art computer vision techniques for various tasks, particularly for image classification. However, there are domains where the training of classification models that generalize on several datasets is still an open challenge because of the highly heterogeneous data and the lack of large datasets with local annotations of the regions of interest, such as histopathology image analysis. Histopathology concerns the microscopic analysis of tissue specimens processed in glass slides to identify diseases such as cancer. Digital pathology concerns the acquisition, management and automatic analysis of digitized histopathology images that are large, having in the order of 100'0002 pixels per image. Digital histopathology images are highly heterogeneous due to the variability of the image acquisition procedures. Creating locally labeled regions (required for the training) is time-consuming and often expensive in the medical field, as physicians usually have to annotate the data. Despite the advances in deep learning, leveraging strongly and weakly annotated datasets to train classification models is still an unsolved problem, mainly when data are very heterogeneous. Large amounts of data are needed to create models that generalize well. This paper presents a novel approach to train CNNs that generalize to heterogeneous datasets originating from various sources and without local annotations. The data analysis pipeline targets Gleason grading on prostate images and includes two models in sequence, following a teacher/student training paradigm. The teacher model (a high-capacity neural network) automatically annotates a set of pseudo-labeled patches used to train the student model (a smaller network). The two models are trained with two different teacher/student approaches: semi-supervised learning and semi-weekly supervised learning. For each of the two approaches, three student training variants are presented. The baseline is provided by training the student model only with the strongly annotated data. Classification performance is evaluated on the student model at the patch level (using the local annotations of the Tissue Micro-Arrays Zurich dataset) and at the global level (using the TCGA-PRAD, The Cancer Genome Atlas-PRostate ADenocarcinoma, whole slide image Gleason score). The teacher/student paradigm allows the models to better generalize on both datasets, despite the inter-dataset heterogeneity and the small number of local annotations used. The classification performance is improved both at the patch-level (up to κ=0.6127±0.0133 from κ=0.5667±0.0285), at the TMA core-level (Gleason score) (up to κ=0.7645±0.0231 from κ=0.7186±0.0306) and at the WSI-level (Gleason score) (up to κ=0.4529±0.0512 from κ=0.2293±0.1350). The results show that with the teacher/student paradigm, it is possible to train models that generalize on datasets from entirely different sources, despite the inter-dataset heterogeneity and the lack of large datasets with local annotations.
Collapse
|