41882-Jo, Taeho
Faculty

Taeho Jo, PhD

Assistant Research Professor of Radiology & Imaging Sciences

Address
GH 4093
RADY
IN
Indianapolis, IN

Bio

Dr. Taeho Jo is a research professional specializing in computational biology and AI/deep learning. He obtained his Ph.D. in a field of Computational Biology from Tokyo Medical and Dental University in 2010. His research leverages large volumes of data and AI/deep learning techniques to explore proteomics, genomics, radiology, and metabolomics, focusing on biomedical science and precision medicine for projects related to Alzheimer's Disease. He is also the renowned author of "Deep Learning for Everyone", a book widely adopted as an official AI education textbook across 28 universities (as of June 2023), providing guidance to both AI learners and practicing computational biologists.

Key Publications

Taeho Jo, Kwangsik Nho*, Paula Bice, and Andrew J. Saykin*. "Deep learning-based identification of genetic variants: Application to Alzheimer’s disease classification." Briefings in Bioinformatics (2022)
Taeho Jo, Kwangsik Nho, Shannon L. Risacher, and Andrew J. Saykin*. "Deep Learning Detection of Informative Features in Tau PET for Alzheimer’s Disease Classification." BMC Bioinformatics (2020)
Taeho Jo*, Kwangsik Nho, and Andrew J. Saykin. "Deep Learning in Alzheimer's disease: Diagnostic Classification and Prognostic Prediction using Neuroimaging Data." Frontiers in Aging Neuroscience (2019) 11:220.

Titles & Appointments

  • Assistant Research Professor of Radiology & Imaging Sciences
  • Education
    2010 PhD Tokyo Medical and Dental University
    2002 BA Inha University
  • Research
    Alzheimer's Disease (AD) is intricately linked with abnormal tau protein accumulation. One of the crucial tasks in AD research is the identification of single nucleotide polymorphisms (SNPs) and associated metabolomics data, as these provide a deeper understanding of the disease's pathogenesis. Dr. Taeho Jo, along with his team, has been striving to devise an innovative deep learning strategy that integrates neuroimaging, genetic data, and metabolomic information, intending to build a more comprehensive understanding of AD.

    Methodological Workflow of the SWAT-ensemble Deep Learning Model


    The image above offers a schematic overview of the SWAT-ensemble deep learning model used in Dr. Jo's study. The workflow involves three primary stages: processing of neuroimaging, genetic, and metabolomics data; application of the convolutional neural network (CNN) and Layer-wise Relevance Propagation (LRP) for feature extraction; and consolidation of the data using the SWAT and c-SWAT methods.

    The SWAT and c-SWAT methods were developed by Dr. Jo and his colleagues and have been detailed in the following publications:

    1. "Deep Learning Detection of Informative Features in Tau PET for Alzheimer’s Disease Classification." BMC Bioinformatics (2020) – This paper discusses the application of deep learning for the detection of informative features in tau PET scans, which is critical for AD classification.

    2. "Deep learning-based identification of genetic variants: Application to Alzheimer’s disease classification." Briefings in Bioinformatics (2022) – In this paper, Dr. Jo and his team present the application of deep learning for the identification of genetic variants that can be used in AD classification.

    3. "Novel circling SWAT for deep learning based diagnostic classification of Alzheimer’s disease: Application to metabolome data." Alzheimer's & Dementia 18 (2022): e069310 – This research introduces the novel circling SWAT method for deep learning-based diagnostic classification of AD, with an emphasis on metabolome data.

  • Professional Organizations
    Alzheimer's Association
    International Society for Computational Biology

Looking for patient care?

To schedule an appointment with a faculty member physician of IU School of Medicine, contact Indiana University Health at 888-484-3258 or use the physician finder by clicking the button below.

Find a doctor