Dr Mason Crossman

Dr Mason Crossman

orcid

Research Focus & Metrics

Academic Career Length

4Years (2023-Present)

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

Representative Publications

View All Publications
6 February 2026 in Radiography
Verified
Journal Article
Open access
MCMinh Chau
Minh Chau
KBKelly Bentley-Spuur
Kelly Bentley-Spuur
SWSamuel White
Samuel White
APAlicia Pyper
Alicia Pyper
DCDr Mason Crossman
Dr Mason Crossman
Artificial intelligence (AI) and machine learning (ML) are increasingly integrated into diagnostic imaging. This review examines how AI adoption affects malpractice risk, the legal standard of care, liability distribution, and informed consent. It also evaluates regulatory developments and ethical concerns, including explicability, autonomy, and professional accountability. AI-supported image interpretation can improve diagnostic accuracy and efficiency. Its integration is reshaping expectations of reasonable clinical practice, with the potential for negligence claims both when clinicians fail to use validated systems and when they rely on insufficiently tested tools. Liability is uncertain because diagnostic responsibility is distributed across clinicians, healthcare organisations, and developers. Existing negligence frameworks assume human reasoning and struggle to accommodate opaque algorithmic decision-making, limiting courts' ability to assess whether AI-assisted diagnoses meet accepted standards. "Black box" models heighten automation bias, hinder legal scrutiny of error, and complicate professional accountability. Informed consent case law suggests AI involvement should be disclosed when it introduces material differences in risk or outcome, although this remains inconsistently applied. Ethical challenges include threats to patient trust, potential clinician deskilling, and reduced transparency in clinical communication. Regulatory initiatives such as the European Union's General Data Protection Regulation and AI Act move toward clearer governance through requirements for data quality, human oversight, and post-market monitoring, yet explicit malpractice guidance remains under-developed globally. Traditional legal and ethical frameworks insufficiently address accountability for AI-driven diagnostic errors. Clarifying responsibility, decision authority, and validation requirements is essential to safeguard patient safety and clinician protection. Clinical protocols should specify approved use cases, oversight expectations, documentation of AI involvement, and management of clinician-algorithm disagreement. Training should support critical review of outputs to mitigate automation bias.

9 May 2026 in Eye
Verified
Journal Article
Open access
PDF file
DCDr Mason Crossman
Dr Mason Crossman
JWJeffrey Weng
Jeffrey Weng
SBStephen Bacchi
Stephen Bacchi
WCWeng Onn Chan
Weng Onn Chan

12 December 2025 in Case Reports in Neurology
Verified
Journal Article
Open access
DCDr Mason Crossman
Dr Mason Crossman
JLJasmin Chau Minh Le
Jasmin Chau Minh Le
AWAlexander William Wright
Alexander William Wright
TGToby Gilbert
Toby Gilbert
JMJohn Maddison
John Maddison
SBStephen Bacchi
Stephen Bacchi
Introduction: Epilepsy surgery remains underutilized despite strong evidence supporting its efficacy for appropriately selected patients with drug-resistant epilepsy (DRE). Case Presentation: We describe the case of a 23-year-old woman with a 4-year history of refractory focal impaired awareness seizures, experiencing up to four episodes per week despite multiple trialed antiseizure medications at tolerated doses. Initial investigations including magnetic resonance imaging (MRI), EEG, and paraneoplastic screening were unremarkable, and she was managed as having non-lesional left temporal lobe epilepsy. As part of an artificial intelligence driven research initiative, a natural language processing based random forest algorithm reviewing neurology clinic documentation identified her as a potential surgical candidate. Complementary large language model analysis supported extraction of relevant seizure and imaging information. This prompted re-evaluation of the case, with PET demonstrating left temporal hypometabolism and repeat MRI revealing a previously unrecognized encephalocele. She subsequently underwent left temporal lobe polectomy with encephalocele disconnection. The procedure was successful, and she has remained seizure-free for 1 year postoperatively, allowing initiation of ASM weaning and the possibility of long-term cure. Conclusion: This case illustrates the transformative potential of epilepsy surgery for individuals with DRE and the promising role of artificial intelligence augmented triage systems in reducing delays to surgical referral. Given the morbidity, mortality, and economic burden associated with DRE, timely surgical evaluation should be considered the standard of care for eligible patients. Integrating artificial intelligence tools into clinical workflows may help overcome longstanding barriers to access and ensure that life-changing interventions are offered earlier in the disease course.

6 May 2026 in Journal of Clinical Epidemiology
Verified
Journal Article
Open access
SWSamuel White
Samuel White
MLMiranda W. Langendam
Miranda W. Langendam
BYBada Yang
Bada Yang
YVYasaman Vali
Yasaman Vali
YCYaxin Chen
Yaxin Chen
HSHolger J. Schünemann
Holger J. Schünemann
The reference standard, defined as the best available test or test strategy for diagnosing a target disease/condition in a specific clinical population, is rarely perfectly accurate. This imperfection raises the question of when (and how) to establish a new test as the reference standard test. The objective of this scoping review was to assess methodological priorities for establishing a new reference standard for medical tests. This scoping review was performed using JBI methodology and reported in accordance with the PRISMA-ScR reporting guideline. We included sources describing methodological approaches to establishing a new reference standard across all healthcare populations and settings. Data were analysed using a combination of descriptive and qualitative content and thematic analyses. We screened 5,396 studies and included 13 in our final analysis (7 methodological studies and 6 applied studies). Three main methodological themes emerged: 1) demonstrating the imperfectness of the current reference standard, 2) comparing candidate reference standard tests, and 3) decision principles for accepting a new reference standard. Amongst the six applied studies that established reference standards for specific conditions, five used composite reference standards. Based on the findings of our review, we define a potential workflow for structuring decision-making relating to whether to accept a new test as the reference standard. Development of formal methodological guidance for when and how to establish or replace a reference standard would likely benefit guideline developers, clinicians, technology assessors and, ultimately, patients.

19 January 2026 in Research Square
Verified
Preprint
Open access
PDF file
DSDarshan Shrestha
Darshan Shrestha
JBJamie Bellinge
Jamie Bellinge
SPSandy Patel
Sandy Patel
RGRudy Goh
Rudy Goh
AWAmy Weber
Amy Weber
SBStephen Bacchi
Stephen Bacchi
Introduction: In embolic stroke of undetermined source (ESUS) 18F-Fluorodeoxyglucose positron emission tomography (18F-FDG-PET) may have a role in detecting several conditions, including malignancies. Currently, there are limited data and no guidelines to inform the use of 18F-FDG-PET in patients with ESUS. This study was conducted with the aim of determining which factors predict a significant abnormality on 18F-FDG-PET in patients with ESUS, to compare with existing literature from other regions. Patients and methods: This cohort study encompassed all public neurology inpatients in South Australia during a two-year period. Patients admitted with ESUS who received an 18F-FDG-PET were included in the study and underwent case note review for clinical, laboratory, and imaging features that may predict significant abnormalities on 18F-FDG-PET. Results: Of 6,786 eligible cases, 36 fulfilled inclusion criteria. In these patients, 8/36 (22.2%) had a significant abnormality on 18F-FDG-PET, all of which were malignancy. These findings informed the stroke aetiology in 7/8 (87.5%) cases. In 7/8 (87.5%) of these cases, the positive 18F-FDG-PET was preceded by alternative imaging evidence of a significant abnormality. The one case without a prior radiological abnormality was due to endometrial cancer. Of those with the three-territory sign, 75% (3/4) had a positive 18F-FDG-PET. Conversely, 90.9% (10/11) of cases with other systemic vascular events had no significant abnormalities on 18F-FDG-PET. Conclusions: In this cohort, nearly all cases with significant abnormalities detected on 18F-FDG-PET in patients with ESUS were preceded by other radiological findings suggestive of a significant abnormality.

5 October 2026 in JEADV Clinical Practice
Verified
Journal Article
Open access
PDF file
DCDr Mason Crossman
Dr Mason Crossman
JWJeffrey Weng
Jeffrey Weng
AVAlysha Vuong
Alysha Vuong
LLLawrence Lin
Lawrence Lin
ZLZhao Feng Liu
Zhao Feng Liu
CCChristopher Y. Chew
Christopher Y. Chew
Background: Subcorneal pustular dermatosis (SPD), or Sneddon–Wilkinson disease, is a rare, relapsing pustular skin condition of unknown cause, characterised by subcorneal pustules. Objectives: To summarise the demographic, clinical, histopathological, management and disease course of SPD. Methods: A systematic review was conducted following PRISMA guidelines including all peer‐reviewed, English‐language studies reporting one or more SPD cases. Results: A total of 169 studies (17 case series, 152 case reports) and 205 patients were identified. The median age at diagnosis was 50 years, with a female‐to‐male ratio of 1.77:1. Over two‐thirds of patients had no comorbidities. Monoclonal gammopathy of unknown significance (MGUS) was the most common comorbidity (9.8%), followed by connective tissue disease (7.3%). Clinical pustules were observed in 92.3% of patients, most commonly affecting the trunk (75.1%), limbs (62%), axilla (44.4%) and groin (36.1%). Pruritus was the most common symptom (44.9%), with pain, burning, or irritation occurring in 16.3%. The median time to diagnosis in cases presenting with rash only was 3 years. A precipitant was identified in 22.9% of cases, including infections, flare of underlying condition, or medications such as calcium channel blockers, TNF‐α inhibitors, and antibiotics. In patients who had histopathological findings reported, subcorneal pustules and perivascular infiltrate most common. Dapsone was the most frequently used treatment (133/189, 70.4%), achieving complete response in 33.8%. Topical and systemic corticosteroids achieved complete responses in 22.5% and 29.6%, respectively. Biologics, mostly TNF‐α inhibitors, were used in 11.6%, with complete and partial responses in 63.6% and 27.3%. Limitations: The review's findings are limited by a limited sample size of heterogeneous, observational data, which can be prone to publication bias, so conclusions on SPD's clinical features and treatment efficacy should be interpreted cautiously pending larger, controlled studies. Conclusions: SPD remains a diagnostic and therapeutic challenge, with frequent delays and limited effectiveness of current treatments. These findings underscore the urgent need for prospective studies and targeted therapies to improve outcomes in this rare dermatosis. Systematic Review Registration: PROSPERO CRD42022344461.