Glioma Image-Level and Slide-Level Gene Predictor (GLISP) for Molecular Diagnosis and Predicting Genetic Events of Adult Diffuse Glioma
The latest World Health Organization (WHO) classification of central nervous system tumors (WHO2021/5th) has incorporated molecular information into the diagnosis of each brain tumor type including diffuse glioma. Therefore, an artificial intelligence (AI) framework for learning histological pattern...
Saved in:
| Main Authors: | , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2024-12-01
|
| Series: | Bioengineering |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2306-5354/12/1/12 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832589083180793856 |
|---|---|
| author | Minh-Khang Le Masataka Kawai Kenta Masui Takashi Komori Takakazu Kawamata Yoshihiro Muragaki Tomohiro Inoue Ippei Tahara Kazunari Kasai Tetsuo Kondo |
| author_facet | Minh-Khang Le Masataka Kawai Kenta Masui Takashi Komori Takakazu Kawamata Yoshihiro Muragaki Tomohiro Inoue Ippei Tahara Kazunari Kasai Tetsuo Kondo |
| author_sort | Minh-Khang Le |
| collection | DOAJ |
| description | The latest World Health Organization (WHO) classification of central nervous system tumors (WHO2021/5th) has incorporated molecular information into the diagnosis of each brain tumor type including diffuse glioma. Therefore, an artificial intelligence (AI) framework for learning histological patterns and predicting important genetic events would be useful for future studies and applications. Using the concept of multiple-instance learning, we developed an AI framework named GLioma Image-level and Slide-level gene Predictor (GLISP) to predict nine genetic abnormalities in hematoxylin and eosin sections: <i>IDH1/2</i>, <i>ATRX</i>, <i>TP53</i> mutations, <i>TERT</i> promoter mutations, <i>CDKN2A/B</i> homozygous deletion (CHD), <i>EGFR</i> amplification (<i>EGFR</i>amp), 7 gain/10 loss (7+/10−), 1p/19q co-deletion, and <i>MGMT</i> promoter methylation. GLISP consists of a pair of patch-level GLISP-P and patient-level GLISP-W models, each pair of which is for a genetic prediction task, providing flexibility in clinical utility. In this study, the Cancer Genome Atlas whole-slide images (WSIs) were used to train the model. A total of 108 WSIs from the Tokyo Women’s Medical University were used as the external dataset. In cross-validation, GLISP yielded patch-level/case-level predictions with top performances in <i>IDH1/2</i> and 1p/19q co-deletion with average areas under the curve (AUCs) of receiver operating characteristics of 0.75/0.79 and 0.73/0.80, respectively. In external validation, the patch-level/case-level AUCs of <i>IDH1/2</i> and 1p/19q co-deletion detection were 0.76/0.83 and 0.78/0.88, respectively. The accuracy in diagnosing IDH-mutant astrocytoma, oligodendroglioma, and IDH-wild-type glioblastoma was 0.66, surpassing the human pathologist average of 0.62 (0.54–0.67). In conclusion, GLISP is a two-stage AI framework for histology-based prediction of genetic events in adult gliomas, which is helpful in providing essential information for WHO 2021 molecular diagnoses. |
| format | Article |
| id | doaj-art-e37b54088f3b4d36b8dc2286d2e18ce9 |
| institution | Kabale University |
| issn | 2306-5354 |
| language | English |
| publishDate | 2024-12-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Bioengineering |
| spelling | doaj-art-e37b54088f3b4d36b8dc2286d2e18ce92025-01-24T13:22:57ZengMDPI AGBioengineering2306-53542024-12-011211210.3390/bioengineering12010012Glioma Image-Level and Slide-Level Gene Predictor (GLISP) for Molecular Diagnosis and Predicting Genetic Events of Adult Diffuse GliomaMinh-Khang Le0Masataka Kawai1Kenta Masui2Takashi Komori3Takakazu Kawamata4Yoshihiro Muragaki5Tomohiro Inoue6Ippei Tahara7Kazunari Kasai8Tetsuo Kondo9Department of Pathology, University of Yamanashi, Yamanashi 409-3898, JapanDepartment of Pathology, University of Yamanashi, Yamanashi 409-3898, JapanDepartment of Pathology, Tokyo Women’s Medical University, Tokyo 162-8666, JapanDepartment of Laboratory Medicine and Pathology (Neuropathology), Tokyo Metropolitan Neurological Hospital, Tokyo 183-0042, JapanDepartment of Neurosurgery, Tokyo Women’s Medical University, Shinjuku, Tokyo 162-8666, JapanCenter for Advanced Medical Engineering Research and Development, Kobe University, Kobe 650-0047, Hyogo, JapanDepartment of Pathology, University of Yamanashi, Yamanashi 409-3898, JapanDepartment of Pathology, University of Yamanashi, Yamanashi 409-3898, JapanDepartment of Pathology, University of Yamanashi, Yamanashi 409-3898, JapanDepartment of Pathology, University of Yamanashi, Yamanashi 409-3898, JapanThe latest World Health Organization (WHO) classification of central nervous system tumors (WHO2021/5th) has incorporated molecular information into the diagnosis of each brain tumor type including diffuse glioma. Therefore, an artificial intelligence (AI) framework for learning histological patterns and predicting important genetic events would be useful for future studies and applications. Using the concept of multiple-instance learning, we developed an AI framework named GLioma Image-level and Slide-level gene Predictor (GLISP) to predict nine genetic abnormalities in hematoxylin and eosin sections: <i>IDH1/2</i>, <i>ATRX</i>, <i>TP53</i> mutations, <i>TERT</i> promoter mutations, <i>CDKN2A/B</i> homozygous deletion (CHD), <i>EGFR</i> amplification (<i>EGFR</i>amp), 7 gain/10 loss (7+/10−), 1p/19q co-deletion, and <i>MGMT</i> promoter methylation. GLISP consists of a pair of patch-level GLISP-P and patient-level GLISP-W models, each pair of which is for a genetic prediction task, providing flexibility in clinical utility. In this study, the Cancer Genome Atlas whole-slide images (WSIs) were used to train the model. A total of 108 WSIs from the Tokyo Women’s Medical University were used as the external dataset. In cross-validation, GLISP yielded patch-level/case-level predictions with top performances in <i>IDH1/2</i> and 1p/19q co-deletion with average areas under the curve (AUCs) of receiver operating characteristics of 0.75/0.79 and 0.73/0.80, respectively. In external validation, the patch-level/case-level AUCs of <i>IDH1/2</i> and 1p/19q co-deletion detection were 0.76/0.83 and 0.78/0.88, respectively. The accuracy in diagnosing IDH-mutant astrocytoma, oligodendroglioma, and IDH-wild-type glioblastoma was 0.66, surpassing the human pathologist average of 0.62 (0.54–0.67). In conclusion, GLISP is a two-stage AI framework for histology-based prediction of genetic events in adult gliomas, which is helpful in providing essential information for WHO 2021 molecular diagnoses.https://www.mdpi.com/2306-5354/12/1/12gliomaartificial intelligencedeep learningnew WHO classificationgenetic abnormalities |
| spellingShingle | Minh-Khang Le Masataka Kawai Kenta Masui Takashi Komori Takakazu Kawamata Yoshihiro Muragaki Tomohiro Inoue Ippei Tahara Kazunari Kasai Tetsuo Kondo Glioma Image-Level and Slide-Level Gene Predictor (GLISP) for Molecular Diagnosis and Predicting Genetic Events of Adult Diffuse Glioma Bioengineering glioma artificial intelligence deep learning new WHO classification genetic abnormalities |
| title | Glioma Image-Level and Slide-Level Gene Predictor (GLISP) for Molecular Diagnosis and Predicting Genetic Events of Adult Diffuse Glioma |
| title_full | Glioma Image-Level and Slide-Level Gene Predictor (GLISP) for Molecular Diagnosis and Predicting Genetic Events of Adult Diffuse Glioma |
| title_fullStr | Glioma Image-Level and Slide-Level Gene Predictor (GLISP) for Molecular Diagnosis and Predicting Genetic Events of Adult Diffuse Glioma |
| title_full_unstemmed | Glioma Image-Level and Slide-Level Gene Predictor (GLISP) for Molecular Diagnosis and Predicting Genetic Events of Adult Diffuse Glioma |
| title_short | Glioma Image-Level and Slide-Level Gene Predictor (GLISP) for Molecular Diagnosis and Predicting Genetic Events of Adult Diffuse Glioma |
| title_sort | glioma image level and slide level gene predictor glisp for molecular diagnosis and predicting genetic events of adult diffuse glioma |
| topic | glioma artificial intelligence deep learning new WHO classification genetic abnormalities |
| url | https://www.mdpi.com/2306-5354/12/1/12 |
| work_keys_str_mv | AT minhkhangle gliomaimagelevelandslidelevelgenepredictorglispformoleculardiagnosisandpredictinggeneticeventsofadultdiffuseglioma AT masatakakawai gliomaimagelevelandslidelevelgenepredictorglispformoleculardiagnosisandpredictinggeneticeventsofadultdiffuseglioma AT kentamasui gliomaimagelevelandslidelevelgenepredictorglispformoleculardiagnosisandpredictinggeneticeventsofadultdiffuseglioma AT takashikomori gliomaimagelevelandslidelevelgenepredictorglispformoleculardiagnosisandpredictinggeneticeventsofadultdiffuseglioma AT takakazukawamata gliomaimagelevelandslidelevelgenepredictorglispformoleculardiagnosisandpredictinggeneticeventsofadultdiffuseglioma AT yoshihiromuragaki gliomaimagelevelandslidelevelgenepredictorglispformoleculardiagnosisandpredictinggeneticeventsofadultdiffuseglioma AT tomohiroinoue gliomaimagelevelandslidelevelgenepredictorglispformoleculardiagnosisandpredictinggeneticeventsofadultdiffuseglioma AT ippeitahara gliomaimagelevelandslidelevelgenepredictorglispformoleculardiagnosisandpredictinggeneticeventsofadultdiffuseglioma AT kazunarikasai gliomaimagelevelandslidelevelgenepredictorglispformoleculardiagnosisandpredictinggeneticeventsofadultdiffuseglioma AT tetsuokondo gliomaimagelevelandslidelevelgenepredictorglispformoleculardiagnosisandpredictinggeneticeventsofadultdiffuseglioma |