Gerald Lip

Gerald Lip

Auto-generated profile, based on publicly available information

Gerald Lip

Breast Unit, Aberdeen Royal Infirmary

Research Focus & Metrics

Academic Career Length

6Years (2021-Present)

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

Representative Publications

View All Publications
1 May 2023 in Radiology: Artificial Intelligence
Verified
Journal Article
Closed access
CVClarisse F. de Vries
Clarisse F. de Vries
SCSamantha J. Colosimo
Samantha J. Colosimo
RSRoger T. Staff
Roger T. Staff
JDJaroslaw A. Dymiter
Jaroslaw A. Dymiter
JYJoseph Yearsley
Joseph Yearsley
CMColin McSkimming
Colin McSkimming
Artificial intelligence (AI) tools may assist breast screening mammography programs, but limited evidence supports their generalizability to new settings. This retrospective study used a 3-year dataset (April 1, 2016–March 31, 2019) from a U.K. regional screening program. The performance of a commercially available breast screening AI algorithm was assessed with a prespecified and site-specific decision threshold to evaluate whether its performance was transferable to a new clinical site. The dataset consisted of women (aged approximately 50–70 years) who attended routine screening, excluding self-referrals, those with complex physical requirements, those who had undergone a previous mastectomy, and those who underwent screening that had technical recalls or did not have the four standard image views. In total, 55 916 screening attendees (mean age, 60 years ± 6 [SD]) met the inclusion criteria. The prespecified threshold resulted in high recall rates (48.3%, 21 929 of 45 444), which reduced to 13.0% (5896 of 45 444) following threshold calibration, closer to the observed service level (5.0%, 2774 of 55 916). Recall rates also increased approximately threefold following a software upgrade on the mammography equipment, requiring per–software version thresholds. Using software-specific thresholds, the AI algorithm would have recalled 277 of 303 (91.4%) screen-detected cancers and 47 of 138 (34.1%) interval cancers. AI performance and thresholds should be validated for new clinical settings before deployment, while quality assurance systems should monitor AI performance for consistency.

1 October 2021 in Cancer Epidemiology
Verified
Journal Article
Closed access
LWLauren T. Walji
Lauren T. Walji
PMPeter Murchie
Peter Murchie
GLGerald Lip
Gerald Lip
VSValerie Speirs
Valerie Speirs
LILisa Iversen
Lisa Iversen

9 December 2022 in Insights into Imaging
Verified
Journal Article
Open access
PDF file
CVClarisse Florence de Vries
Clarisse Florence de Vries
SCSamantha J. Colosimo
Samantha J. Colosimo
MBMoragh Boyle
Moragh Boyle
GLGerald Lip
Gerald Lip
LALesley A. Anderson
Lesley A. Anderson
CMC. McSkimming
C. McSkimming
Objectives: This study surveyed the views of breast screening readers in the UK on how to incorporate Artificial Intelligence (AI) technology into breast screening mammography. Methods: An online questionnaire was circulated to the UK breast screening readers. Questions included their degree of approval of four AI implementation scenarios: AI as triage, AI as a companion reader/reader aid, AI replacing one of the initial two readers, and AI replacing all readers. They were also asked to rank five AI representation options (discrete opinion; mammographic scoring; percentage score with 100% indicating malignancy; region of suspicion; heat map) and indicate which evidence they considered necessary to support the implementation of AI into their practice among six options offered. Results: The survey had 87 nationally accredited respondents across the UK; 73 completed the survey in full. Respondents approved of AI replacing one of the initial two human readers and objected to AI replacing all human readers. Participants were divided on AI as triage and AI as a reader companion. A region of suspicion superimposed on the image was the preferred AI representation option. Most screen readers considered national guidelines (77%), studies using a nationally representative dataset (65%) and independent prospective studies (60%) as essential evidence. Participants’ free-text comments highlighted concerns and the need for additional validation. Conclusions: Overall, screen readers supported the introduction of AI as a partial replacement of human readers and preferred a graphical indication of the suspected tumour area, with further evidence and national guidelines considered crucial prior to implementation.

10 March 2026 in Nature Cancer
Verified
Journal Article
Open access
PDF file
CVClarisse Florence de Vries
Clarisse Florence de Vries
GLGerald Lip
Gerald Lip
RSRoger Todd Staff
Roger Todd Staff
JDJaroslaw Artur Dymiter
Jaroslaw Artur Dymiter
BTBenjamin Tse
Benjamin Tse
LALesley Ann Anderson
Lesley Ann Anderson
Artificial intelligence (AI) tools can improve breast screening performance but different screening sites have varying needs. Here the GEMINI prospective evaluation of 10,889 women, within one UK region, used both live AI integration and simulations to model 17 different ways AI could be used in breast screening. All women received routine care. One AI tool was assessed. When the AI tool recommended recall but routine double reading did not, cases underwent additional human review, detecting 11 additional cancers. The primary AI workflow could improve cancer detection by 10.4% (1 per 1,000), maintain the recall rate (0.8% reduction) and reduce workload by up to 31%. Other workflow variations significantly improved all measured metrics (superiority in cancer detection rate, recall rate, positive predictive value (PPV), sensitivity and specificity) with up to 36% workload savings. Different AI integrations in breast screening could offer various clinical and operational gains, allowing for adaptation to local healthcare needs.

30 October 2024 in Communications Medicine
Verified
Journal Article
Open access
PDF file
VMVasiliki Mallikourti
Vasiliki Mallikourti
PRP. James Ross
P. James Ross
OMOliver Maier
Oliver Maier
KHKatie Hanna
Katie Hanna
EHEhab Husain
Ehab Husain
LBLionel Broche
Lionel Broche
Background: This prospective feasibility study explores Field-Cycling Imaging (FCI), a new MRI technology that measures the longitudinal relaxation time across a range of low magnetic field strengths, providing additional information about the molecular properties of tissues. This study aims to assess the performance of FCI and investigate new quantitative biomarkers at low fields within the context of breast cancer. Methods: We conducted a study involving 9 people living with breast cancer (10 tumours in total, mean age, 54 ± 10 years). FCI images were obtained at four magnetic field strengths (2.3 mT to 200 mT). FCI images were processed to generate T1 maps and 1/T1 dispersion profiles from regions of tumour, normal adipose tissue, and glandular tissue. The dispersion profiles were subsequently fitted using a power law model. Statistical analysis focused on comparing potential FCI biomarkers using a Mann-Whitney U or Wilcoxon signed rank test. Results: We show that low magnetic fields clearly differentiate tumours from adipose and glandular tissues without contrast agents, particularly at 22 mT (1/T1, median [IQR]: 6.8 [3.9–7.8] s−1 vs 9.1 [8.9–10.2] s−1 vs 8.1 [6.2–9.2] s−1, P < 0.01), where the tumour-to-background contrast ratio was highest (62%). Additionally, 1/T1 dispersion indicated a potential to discriminate invasive from non-invasive cancers (median [IQR]: 0.05 [0.03–0.09] vs 0.19 [0.09–0.26], P = 0.038). Conclusions: To the best of our knowledge, we described the first application of in vivo FCI in breast cancer, demonstrating relevant biomarkers that could complement diagnosis of current imaging modalities, non-invasively and without contrast agents.

1 January 2024 in BJR|Artificial Intelligence
Verified
Journal Article
Open access
PDF file
GLGerald Lip
Gerald Lip
ANAlex Novak
Alex Novak
MGMathias Goyen
Mathias Goyen
KBKatherine Boylan
Katherine Boylan
AKAmrita Kumar
Amrita Kumar
Clinical care, workflow, and patient outcomes within National Health Service (NHS) radiology services will potentially benefit significantly in coming years with the integration of artificial intelligence (AI) into various specialty areas. This article summarizes an expert roundtable discussion, convened by the British Institute of Radiology, comprised experts from clinical, NHS management and industry backgrounds. Facilitators and barriers to both implementation and integration of AI platforms into daily practice are discussed alongside examples including AI in emergency care and breast screening. Issues addressed include the need to generate more real-world evidence that helps developers and service providers to determine the optimal way for patients to reap the benefits of AI with faster reporting turnaround time, faster diagnoses, and treatment times, for example. To ensure the effective and timely wide-scale implementation, adoption requires a national policy drive and national level coordination but specific, local needs of NHS Trusts and any particular service within Trusts will also need to be met. By embedding AI-capabilities into scanners, ultrasound-devices, X-ray equipment, and Picture archiving and communication system (PACS), local services will benefit, while AI applications can streamline workflows across and between departments and potentially Trusts. AI predictive analytics and systems engineering can be used to better manage patient volume and movement. In the short term, education, training, and confidence in adapting and adopting AI in radiology require dedicated local, national, and governmental effort but the future benefits promise to justify the means.