Chun Fung Kwok

Dr. Chun Fung Kwok


Research Focus & Metrics

Academic Career Length

16Years (2011-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
21 August 2023 in IEEE Transactions on Medical Imaging
Verified
Journal Article
Closed access
CWChong Wang
Chong Wang
YCYuanhong Chen
Yuanhong Chen
FLFengbei Liu
Fengbei Liu
MEMichael Elliott
Michael Elliott
CKChun Fung Kwok
Chun Fung Kwok
GCGustavo Carneiro
Gustavo Carneiro
The deployment of automated deep-learning classifiers in clinical practice has the potential to streamline the diagnosis process and improve the diagnosis accuracy, but the acceptance of those classifiers relies on both their accuracy and interpretability. In general, accurate deep-learning classifiers provide little model interpretability, while interpretable models do not have competitive classification accuracy. In this paper, we introduce a new deep-learning diagnosis framework, called InterNRL, that is designed to be highly accurate and interpretable. InterNRL consists of a student-teacher framework, where the student model is an interpretable prototype-based classifier (ProtoPNet) and the teacher is an accurate global image classifier (GlobalNet). The two classifiers are mutually optimised with a novel reciprocal learning paradigm in which the student ProtoPNet learns from optimal pseudo labels produced by the teacher GlobalNet, while GlobalNet learns from ProtoPNet’s classification performance and pseudo labels. This reciprocal learning paradigm enables InterNRL to be flexibly optimised under both fully- and semi-supervised learning scenarios, reaching state-of-the-art classification performance in both scenarios for the tasks of breast cancer and retinal disease diagnosis. Moreover, relying on weakly-labelled training images, InterNRL also achieves superior breast cancer localisation and brain tumour segmentation results than other competing methods.

30 August 2024 in Nature Communications
Verified
Journal Article
Open access
PDF file
HFHelen M. L. Frazer
Helen M. L. Frazer
CPCarlos A. Peña-Solorzano
Carlos A. Peña-Solorzano
CKChun Fung Kwok
Chun Fung Kwok
MEMichael S. Elliott
Michael S. Elliott
YCYuanhong Chen
Yuanhong Chen
DMDavis J. McCarthy
Davis J. McCarthy
Artificial intelligence (AI) readers of mammograms compare favourably to individual radiologists in detecting breast cancer. However, AI readers cannot perform at the level of multi-reader systems used by screening programs in countries such as Australia, Sweden, and the UK. Therefore, implementation demands human-AI collaboration. Here, we use a large, high-quality retrospective mammography dataset from Victoria, Australia to conduct detailed simulations of five potential AI-integrated screening pathways, and examine human-AI interaction effects to explore automation bias. Operating an AI reader as a second reader or as a high confidence filter improves current screening outcomes by 1.9–2.5% in sensitivity and up to 0.6% in specificity, achieving 4.6–10.9% reduction in assessments and 48–80.7% reduction in human reads. Automation bias degrades performance in multi-reader settings but improves it for single-readers. This study provides insight into feasible approaches for AI-integrated screening pathways and prospective studies necessary prior to clinical adoption.

1 March 2023 in Radiology: Artificial Intelligence
Verified
Journal Article
Open access
HFHelen M. L. Frazer
Helen M. L. Frazer
JTJennifer S. N. Tang
Jennifer S. N. Tang
MEMichael S. Elliott
Michael S. Elliott
KKKatrina M. Kunicki
Katrina M. Kunicki
BHBrendan Hill
Brendan Hill
DMDavis J. McCarthy
Davis J. McCarthy
“Just Accepted” papers have undergone full peer review and have been accepted for publication in Radiology: Artificial Intelligence. This article will undergo copyediting, layout, and proof review ...

10 May 2024 in Medical Image Analysis
Verified
Journal Article
Closed access
YCYuanhong Chen
Yuanhong Chen
YLYuyuan Liu
Yuyuan Liu
CWChong Wang
Chong Wang
MEMichael Elliott
Michael Elliott
CKChun Fung Kwok
Chun Fung Kwok
GCGustavo Carneiro
Gustavo Carneiro

17 January 2023 in Atmosphere
Verified
Journal Article
Open access
PDF file
CKChun Fung Kwok
Chun Fung Kwok
GQGuoqi Qian
Guoqi Qian
YKYuriy Kuleshov
Yuriy Kuleshov
In this paper, we study the problem of extracting trends from time series data involving missing values. In particular, we investigate a general class of procedures that impute the missing data and then extract trends using seasonal-trend decomposition based on loess (STL), where loess stands for locally weighted smoothing, a popular tool for describing the regression relationship between two variables by a smooth curve. We refer to them as the imputation-STL procedures. Two results are obtained in this paper. First, we settle a theoretical issue, namely the connection between imputation error and the overall error from estimating the trend. Specifically, we derive the bounds for the overall error in terms of the imputation error. This subsequently facilitates the error analysis of any imputation-STL procedure and justifies its use in practice. Second, we investigate loess-STL, a particular imputation-STL procedure with the imputation also being performed using loess. Through both theoretical arguments and simulation results, we show that loess-STL has the capacity of handling a high proportion of missing data and providing reliable trend estimates if the underlying trend is smooth and the missing data are dispersed over the time series. In addition to mathematical derivations and simulation study, we apply our loess-STL procedure to profile radiosonde records of upper air temperature at 22 Antarctic research stations covering the past 50 years. For purpose of illustration, we present in this paper only the results for Novolazaravskaja station which has temperature records with more than 8.4% dispersed missing values at 8 pressure levels from October/1969 to March/2011.

3 March 2026 in The Lancet Digital Health
Verified
Journal Article
Open access
HFHelen M L Frazer
Helen M L Frazer
JHJohn Hopper
John Hopper
TNTuong L. Nguyen
Tuong L. Nguyen
MEMichael S Elliott
Michael S Elliott
KKKatrina M Kunicki
Katrina M Kunicki
DMDavis J McCarthy
Davis J McCarthy
Artificial intelligence (AI)-based algorithms are being implemented in breast screening to detect breast cancers on mammographic images. We aimed to apply an epidemiological approach to demonstrate how a cancer detection algorithm can be leveraged as an intermediate-term predictor of breast cancer (current and 4-year risk) to deliver greater risk-based personalisation in screening mammography. In this population cohort study, we used detection scores from an AI cancer detection algorithm (BRAIx AI Reader), which was calibrated using a training dataset of 397 648 women aged 40 years to 97 years from women who screened at BreastScreen Victoria, Australia between Jan 1, 2016, and Dec 31, 2017, to create a woman-specific mammography-based score for breast cancer risk, the BRAIx risk score. Subsequently, the BRAIx risk score was evaluated on an independent test dataset of women from BreastScreen Victoria, Australia, comprising a random population cohort of 96 348 women who screened from Jan 1, 2016, to Dec 31, 2017, aged 40 years to 74 years, and an independent, external dataset from woman screened at Karolinska University Hospital, Stockholm, Sweden. We applied logistic regression, using the BRAIx risk score to estimate risks of invasive breast cancers on the test dataset: (1) detected at cohort entry (n=525); and (2) for women given an all clear, diagnosed during the next 4 years either at future screens (n=790) or during intervals between screens (n=308). We also trained full multivariate risk models (logistic regression and elastic net) using the training dataset and evaluated their predictive performance on the test and external validation data, with assessment of familial aspects of the BRAIx risk score achieved with inference about causation from examining changes in regression coefficients in an innovative statistical analysis framework. In both Australian and Swedish test datasets, the BRAIx risk score predicted cancer detection at cohort entry and future cancer risk (all p<0·0001). The BRAIx risk score was the strongest tested explanatory factor for cancer detection at cohort entry (odds ratio 13·80 [95% CI 9·54-20·80] in Australian data; 8·89 [3·19-37·49] in Swedish data) and for intermediate-term cancer risk (2·29 [2·13-2.47] in Australian data; 2·15 [1·85-2·50] in Swedish data). We found that adding a thresholded binary version of the BRAIx risk score significantly improved model fit (p<2·2 × 10-16, Australian and Swedish data) and women with BRAIx risk scores of more than 2 were significantly at many-fold increased risk of intermediate-term cancer than women below that threshold (12·34 [7·33-20·91], Australia; 44·7 [11·9-184·9], Sweden; p<0·0001). For the top 2% of women given an all clear with the highest BRAIx risk score, the probability of a cancer diagnosis within 4 years was 9·7%. The BRAIx risk score explained 23% of why family history predicts 4-year risk (p0·05) and became associated with lower risk for intermediate-term cancer in the external Swedish test dataset (0·83 [0·73-0·95]). The BRAIx risk score is a strong intermediate-term predictor of breast cancer (current to 4-year risk). Calibrating the score on a training dataset produces population-specific probabilities for calculating individual-specific risk scores for screening clients based on their mammogram images. These risk scores enable future development of personalised screening pathways to transform population breast cancer screening and save lives. Identification of women given an all clear but at very high risk, similar to those carrying BRCA1 and BRCA2 mutations, could reveal insights into both familial and non-familial causes of breast cancer. Australian Government Medical Research Future Fund, the Ramaciotti Foundation, the National Breast Cancer Foundation, Cancer Australia, and the National Health and Medical Research Council.