Use of federated learning to develop an artificial intelligence model predicting usable blastocyst formation from pre-ICSI oocyte images.

RESEARCH QUESTION: Can federated learning be used to develop an artificial intelligence (AI) model for evaluating oocyte competence using two-dimensional images of denuded oocytes in metaphase II prior to intracytoplasmic sperm injection (ICSI)? RESULTS: The oocyte AI model demonstrated area under the curve (AUC) up to 0.65 on two blind test datasets. High sensitivity for predicting competent oocytes (83-88%) was offset by lower specificity (26-36%). Exclusion of confounding biological variables (male factor infertility and maternal age ≥35 years) improved AUC up to 14%, primarily due to increased specificity. AI score correlated with size of the zona pellucida and perivitelline space, and ooplasm appearance. AI score also correlated with blastocyst expansion grade and morphological quality. The sum of AI scores from oocytes in group culture images predicted the formation of two or more usable blastocysts (AUC 0.77). CONCLUSION: An AI model to evaluate oocyte competence was developed using federated learning, representing an essential step in protecting patient data. The AI model was significantly predictive of oocyte competence, as defined by usable blastocyst formation, which is a critical factor for IVF success. Potential clinical utility ranges from selective oocyte fertilization to guiding treatment decisions regarding additional rounds of oocyte retrieval. DESIGN: In total, 10,677 oocyte images with associated metadata were collected prospectively by eight IVF clinics across six countries. AI training used federated learning, where data were retained on regional servers to comply with data privacy laws. The final AI model required a single image as input to evaluate oocyte competence, which was defined by the formation of a usable blastocyst (≥expansion grade 3 by day 5 or 6 post ICSI).

Development of an artificial intelligence model for predicting the likelihood of human embryo euploidy based on blastocyst images from multiple imaging systems during IVF.

STUDY QUESTION: Can an artificial intelligence (AI) model predict human embryo ploidy status using static images captured by optical light microscopy? SUMMARY ANSWER: Results demonstrated predictive accuracy for embryo euploidy and showed a significant correlation between AI score and euploidy rate, based on assessment of images of blastocysts at Day 5 after IVF. WHAT IS KNOWN ALREADY: Euploid embryos displaying the normal human chromosomal complement of 46 chromosomes are preferentially selected for transfer over aneuploid embryos (abnormal complement), as they are associated with improved clinical outcomes. Currently, evaluation of embryo genetic status is most commonly performed by preimplantation genetic testing for aneuploidy (PGT-A), which involves embryo biopsy and genetic testing. The potential for embryo damage during biopsy, and the non-uniform nature of aneuploid cells in mosaic embryos, has prompted investigation of additional, non-invasive, whole embryo methods for evaluation of embryo genetic status. STUDY DESIGN, SIZE, DURATION: A total of 15 192 blastocyst-stage embryo images with associated clinical outcomes were provided by 10 different IVF clinics in the USA, India, Spain and Malaysia. The majority of data were retrospective, with two additional prospectively collected blind datasets provided by IVF clinics using the genetics AI model in clinical practice. Of these images, a total of 5050 images of embryos on Day 5 of in vitro culture were used for the development of the AI model. These Day 5 images were provided for 2438 consecutively treated women who had undergone IVF procedures in the USA between 2011 and 2020. The remaining images were used for evaluation of performance in different settings, or otherwise excluded for not matching the inclusion criteria. PARTICIPANTS/MATERIALS, SETTING, METHODS: The genetics AI model was trained using static 2-dimensional optical light microscope images of Day 5 blastocysts with linked genetic metadata obtained from PGT-A. The endpoint was ploidy status (euploid or aneuploid) based on PGT-A results. Predictive accuracy was determined by evaluating sensitivity (correct prediction of euploid), specificity (correct prediction of aneuploid) and overall accuracy. The Matthew correlation coefficient and receiver-operating characteristic curves and precision-recall curves (including AUC values), were also determined. Performance was also evaluated using correlation analyses and simulated cohort studies to evaluate ranking ability for euploid enrichment. MAIN RESULTS AND THE ROLE OF CHANCE: Overall accuracy for the prediction of euploidy on a blind test dataset was 65.3%, with a sensitivity of 74.6%. When the blind test dataset was cleansed of poor quality and mislabeled images, overall accuracy increased to 77.4%. This performance may be relevant to clinical situations where confounding factors, such as variability in PGT-A testing, have been accounted for. There was a significant positive correlation between AI score and the proportion of euploid embryos, with very high scoring embryos (9.0-10.0) twice as likely to be euploid than the lowest-scoring embryos (0.0-2.4). When using the genetics AI model to rank embryos in a cohort, the probability of the top-ranked embryo being euploid was 82.4%, which was 26.4% more effective than using random ranking, and ∼13-19% more effective than using the Gardner score. The probability increased to 97.0% when considering the likelihood of one of the top two ranked embryos being euploid, and the probability of both top two ranked embryos being euploid was 66.4%. Additional analyses showed that the AI model generalized well to different patient demographics and could also be used for the evaluation of Day 6 embryos and for images taken using multiple time-lapse systems. Results suggested that the AI model could potentially be used to differentiate mosaic embryos based on the level of mosaicism. LIMITATIONS, REASONS FOR CAUTION: While the current investigation was performed using both retrospectively and prospectively collected data, it will be important to continue to evaluate real-world use of the genetics AI model. The endpoint described was euploidy based on the clinical outcome of PGT-A results only, so predictive accuracy for genetic status in utero or at birth was not evaluated. Rebiopsy studies of embryos using a range of PGT-A methods indicated a degree of variability in PGT-A results, which must be considered when interpreting the performance of the AI model. WIDER IMPLICATIONS OF THE FINDINGS: These findings collectively support the use of this genetics AI model for the evaluation of embryo ploidy status in a clinical setting. Results can be used to aid in prioritizing and enriching for embryos that are likely to be euploid for multiple clinical purposes, including selection for transfer in the absence of alternative genetic testing methods, selection for cryopreservation for future use or selection for further confirmatory PGT-A testing, as required. STUDY FUNDING/COMPETING INTEREST(S): Life Whisperer Diagnostics is a wholly owned subsidiary of the parent company, Presagen Holdings Pty Ltd. Funding for the study was provided by Presagen with grant funding received from the South Australian Government: Research, Commercialisation, and Startup Fund (RCSF). 'In kind' support and embryology expertise to guide algorithm development were provided by Ovation Fertility. 'In kind' support in terms of computational resources provided through the Amazon Web Services (AWS) Activate Program. J.M.M.H., D.P. and M.P. are co-owners of Life Whisperer and Presagen. S.M.D., M.A.D. and T.V.N. are employees or former employees of Life Whisperer. S.M.D, J.M.M.H, M.A.D, T.V.N., D.P. and M.P. are listed as inventors of patents relating to this work, and also have stock options in the parent company Presagen. M.V. sits on the advisory board for the global distributor of the technology described in this study and also received support for attending meetings. TRIAL REGISTRATION NUMBER: N/A.

Development of an artificial intelligence-based assessment model for prediction of embryo viability using static images captured by optical light microscopy during IVF.

STUDY QUESTION: Can an artificial intelligence (AI)-based model predict human embryo viability using images captured by optical light microscopy? SUMMARY ANSWER: We have combined computer vision image processing methods and deep learning techniques to create the non-invasive Life Whisperer AI model for robust prediction of embryo viability, as measured by clinical pregnancy outcome, using single static images of Day 5 blastocysts obtained from standard optical light microscope systems. WHAT IS KNOWN ALREADY: Embryo selection following IVF is a critical factor in determining the success of ensuing pregnancy. Traditional morphokinetic grading by trained embryologists can be subjective and variable, and other complementary techniques, such as time-lapse imaging, require costly equipment and have not reliably demonstrated predictive ability for the endpoint of clinical pregnancy. AI methods are being investigated as a promising means for improving embryo selection and predicting implantation and pregnancy outcomes. STUDY DESIGN, SIZE, DURATION: These studies involved analysis of retrospectively collected data including standard optical light microscope images and clinical outcomes of 8886 embryos from 11 different IVF clinics, across three different countries, between 2011 and 2018. PARTICIPANTS/MATERIALS, SETTING, METHODS: The AI-based model was trained using static two-dimensional optical light microscope images with known clinical pregnancy outcome as measured by fetal heartbeat to provide a confidence score for prediction of pregnancy. Predictive accuracy was determined by evaluating sensitivity, specificity and overall weighted accuracy, and was visualized using histograms of the distributions of predictions. Comparison to embryologists' predictive accuracy was performed using a binary classification approach and a 5-band ranking comparison. MAIN RESULTS AND THE ROLE OF CHANCE: The Life Whisperer AI model showed a sensitivity of 70.1% for viable embryos while maintaining a specificity of 60.5% for non-viable embryos across three independent blind test sets from different clinics. The weighted overall accuracy in each blind test set was >63%, with a combined accuracy of 64.3% across both viable and non-viable embryos, demonstrating model robustness and generalizability beyond the result expected from chance. Distributions of predictions showed clear separation of correctly and incorrectly classified embryos. Binary comparison of viable/non-viable embryo classification demonstrated an improvement of 24.7% over embryologists' accuracy (P = 0.047, n = 2, Student's t test), and 5-band ranking comparison demonstrated an improvement of 42.0% over embryologists (P = 0.028, n = 2, Student's t test). LIMITATIONS, REASONS FOR CAUTION: The AI model developed here is limited to analysis of Day 5 embryos; therefore, further evaluation or modification of the model is needed to incorporate information from different time points. The endpoint described is clinical pregnancy as measured by fetal heartbeat, and this does not indicate the probability of live birth. The current investigation was performed with retrospectively collected data, and hence it will be of importance to collect data prospectively to assess real-world use of the AI model. WIDER IMPLICATIONS OF THE FINDINGS: These studies demonstrated an improved predictive ability for evaluation of embryo viability when compared with embryologists' traditional morphokinetic grading methods. The superior accuracy of the Life Whisperer AI model could lead to improved pregnancy success rates in IVF when used in a clinical setting. It could also potentially assist in standardization of embryo selection methods across multiple clinical environments, while eliminating the need for complex time-lapse imaging equipment. Finally, the cloud-based software application used to apply the Life Whisperer AI model in clinical practice makes it broadly applicable and globally scalable to IVF clinics worldwide. STUDY FUNDING/COMPETING INTEREST(S): Life Whisperer Diagnostics, Pty Ltd is a wholly owned subsidiary of the parent company, Presagen Pty Ltd. Funding for the study was provided by Presagen with grant funding received from the South Australian Government: Research, Commercialisation and Startup Fund (RCSF). 'In kind' support and embryology expertise to guide algorithm development were provided by Ovation Fertility. J.M.M.H., D.P. and M.P. are co-owners of Life Whisperer and Presagen. Presagen has filed a provisional patent for the technology described in this manuscript (52985P pending). A.P.M. owns stock in Life Whisperer, and S.M.D., A.J., T.N. and A.P.M. are employees of Life Whisperer.

HbA1c determination from HemaSpot™ blood collection devices: comparison of home prepared dried blood spots with standard venous blood analysis.

AIM: To assess the clinical performance and patient acceptance of HemaSpot™ blood collection devices as an alternative blood collection method. METHODS: Adult men and women with any type of diabetes, routinely carrying out self-monitoring of blood glucose were recruited (n = 128). Participants provided a venous blood sample and prepared two HemaSpot dried blood spots, one at clinics and one at home. HbA1c analysis was by Tosoh G8 high-performance liquid chromatography. Participants also completed a questionnaire. RESULTS: Strong linear relationships been HbA1c levels in dried blood spots and venous blood were observed and a linear model was fitted to the data. Time between dried blood spot preparation and testing did not impact the model. Participants were accepting of the approach: 69.2% would use this system if available and 60.7% would be more likely to use this system than going to their general practitioner. CONCLUSIONS: The combination of a robust desiccating dried blood spot device, home sample preparation and return by post produces HbA1c data that support the use of a time-independent linear calibration of dried blood spot to venous blood HbA1c . A robust remote sample collection service would be valuable to people living with diabetes in urban areas who are working or house-bound as well as those living in remote or rural locations.

Optimization of whispering gallery resonator design for biosensing applications.

Whispering gallery modes (WGMs) within microsphere cavities enable highly sensitive label-free detection of changes in the surrounding refractive index. This detection modality is of particular interest for biosensing applications. However, the majority of biosensing work utilizing WGMs to date has been conducted with resonators made from either silica or polystyrene, while other materials remain largely uninvestigated. By considering characteristics such as the quality factor and sensitivity of the resonator, the optimal WGM sensor design can be identified for various applications. This work explores the choice of resonator refractive index and size to provide design guidelines for undertaking refractive index biosensing using WGMs.

Efficient automated error detection in medical data using deep-learning and label-clustering.

Medical datasets inherently contain errors from subjective or inaccurate test results, or from confounding biological complexities. It is difficult for medical experts to detect these elusive errors manually, due to lack of contextual information, limiting data privacy regulations, and the sheer scale of data to be reviewed. Current methods for training robust artificial intelligence (AI) models on data containing mislabeled examples generally fall into one of several categories-attempting to improve the robustness of the model architecture, the regularization techniques used, the loss function used during training, or selecting a subset of data that contains cleaner labels. This last category requires the ability to efficiently detect errors either prior to or during training, either relabeling them or removing them completely. More recent progress in error detection has focused on using multi-network learning to minimize deleterious effects of errors on training, however, using many neural networks to reach a consensus on which data should be removed can be computationally intensive and inefficient. In this work, a deep-learning based algorithm was used in conjunction with a label-clustering approach to automate error detection. For dataset with synthetic label flips added, these errors were identified with an accuracy of up to 85%, while requiring up to 93% less computing resources to complete compared to a previous model consensus approach developed previously. The resulting trained AI models exhibited greater training stability and up to a 45% improvement in accuracy, from 69 to over 99% compared to the consensus approach, at least 10% improvement on using noise-robust loss functions in a binary classification problem, and a 51% improvement for multi-class classification. These results indicate that practical, automated a priori detection of errors in medical data is possible, without human oversight.

A novel decentralized federated learning approach to train on globally distributed, poor quality, and protected private medical data.

Training on multiple diverse data sources is critical to ensure unbiased and generalizable AI. In healthcare, data privacy laws prohibit data from being moved outside the country of origin, preventing global medical datasets being centralized for AI training. Data-centric, cross-silo federated learning represents a pathway forward for training on distributed medical datasets. Existing approaches typically require updates to a training model to be transferred to a central server, potentially breaching data privacy laws unless the updates are sufficiently disguised or abstracted to prevent reconstruction of the dataset. Here we present a completely decentralized federated learning approach, using knowledge distillation, ensuring data privacy and protection. Each node operates independently without needing to access external data. AI accuracy using this approach is found to be comparable to centralized training, and when nodes comprise poor-quality data, which is common in healthcare, AI accuracy can exceed the performance of traditional centralized training.

The 40Ar(d,p)41Ar cross section between 3-7 MeV.

To determine the safety of using argon as a deuteron beam stopping material, the  40Ar(d,p)41Ar cross section was measured at average deuteron energies of 3.6 MeV, 5.5 MeV, and 7.0 MeV using an activation method. A 16-MeV deuteron beam produced by Lawrence Berkeley National Laboratory's 88-Inch Cyclotron was degraded to each energy by nickel foils and the front wall of an aluminum gas chamber. The reduced-energy deuterons were used to activate a sample of natAr gas. After each irradiation, the gas chamber's  41Ar activation was measured with a high-purity germanium detector. The cross sections measured were larger than a previous measurement by ∼40%.

Automated detection of poor-quality data: case studies in healthcare.

The detection and removal of poor-quality data in a training set is crucial to achieve high-performing AI models. In healthcare, data can be inherently poor-quality due to uncertainty or subjectivity, but as is often the case, the requirement for data privacy restricts AI practitioners from accessing raw training data, meaning manual visual verification of private patient data is not possible. Here we describe a novel method for automated identification of poor-quality data, called Untrainable Data Cleansing. This method is shown to have numerous benefits including protection of private patient data; improvement in AI generalizability; reduction in time, cost, and data needed for training; all while offering a truer reporting of AI performance itself. Additionally, results show that Untrainable Data Cleansing could be useful as a triage tool to identify difficult clinical cases that may warrant in-depth evaluation or additional testing to support a diagnosis.

Deep crustal drilling in the north atlantic ocean.

Oceanic crustal drilling by R. V. Glomar Challenger at 15 sites in the North Atlantic has led to a complex picture of the upper half kilometer of the crust. Elements of the picture include the absence of the source for linear magnetic anomalies, marked episodicity of volcanic activity, ubiquitous low temperature alteration and evidence for large scale tectonic disturbance. Comparison sections in the Pacific and much deeper crustal drilling are needed to attack problems arising from the North Atlantic results.

Linkage of early-onset familial breast cancer to chromosome 17q21.

Human breast cancer is usually caused by genetic alterations of somatic cells of the breast, but occasionally, susceptibility to the disease is inherited. Mapping the genes responsible for inherited breast cancer may also allow the identification of early lesions that are critical for the development of breast cancer in the general population. Chromosome 17q21 appears to be the locale of a gene for inherited susceptibility to breast cancer in families with early-onset disease. Genetic analysis yields a lod score (logarithm of the likelihood ratio for linkage) of 5.98 for linkage of breast cancer susceptibility to D17S74 in early-onset families and negative lod scores in families with late-onset disease. Likelihood ratios in favor of linkage heterogeneity among families ranged between 2000:1 and greater than 10(6):1 on the basis of multipoint analysis of four loci in the region.

A comparison of the quality of care provided to cancer patients in the UK in the last three months of life in in-patient hospices compared with hospitals, from the perspective of bereaved relatives: results from a survey using the VOICES questionnaire.

The importance of evaluating systematically the effectiveness of hospice care has been noted for at least 20 years. There is, however, limited evidence about whether and how the care provided to terminally ill patients by in-patient hospices in the UK differs from that provided in NHS hospitals. In this article, we, therefore, present a comparison of hospice in-patient care and hospital care for cancer patients in the UK, from the perspective of bereaved relatives who had experienced both types of care during the last 3 months of the patient's life. The Office of National Statistics drew a random sample of 800 deaths in South London in 2002, and sent the person who registered the death (the informant) a Views of Informal Carers - Evaluation of Services (VOICES) questionnaire 3-9 months after the death, with up to two reminders. There was a response rate of 48%. For this analysis, 40 cancer patients whose informant reported both a hospice in-patient admission and a hospital admission in the last 3 months of life were identified. Informants answered the same questions about each admission and responses on these were compared. There were statistically significant differences between respondents' views of hospice and hospital care on eight out of 13 variables measuring aspects of satisfaction with care, with a trend towards statistical significance on a further two: in all cases respondents rated hospice care more positively than hospital care. There were no differences in the experience of pain and breathlessness in the two settings, but respondents rated pain control by the hospice as more effective. In comparison to hospital care, from the perspective of bereaved relatives, hospice in-patient care provided better pain control, better communication with patients and families, and better medical, nursing and personal care, which treated the patient with more dignity. Further research is needed to confirm these findings using a wider sample of in-patient hospices in the UK and including the perspectives of patients. Providing high quality care for terminally ill patients in acute hospitals remains an important challenge.

Author Correction: Rapid susceptibility profiling of carbapenem-resistant Klebsiella pneumoniae.

A correction to this article has been published and is linked from the HTML and PDF versions of this paper. The error has not been fixed in the paper.

Rapid susceptibility profiling of carbapenem-resistant Klebsiella pneumoniae.

The expanding global distribution of multi-resistant Klebsiella pneumoniae demands faster antimicrobial susceptibility testing (AST) to guide antibiotic treatment. Current ASTs rely on time-consuming differentiation of resistance and susceptibility after initial isolation of bacteria from a clinical specimen. Here we describe a flow cytometry workflow to determine carbapenem susceptibility from bacterial cell characteristics in an international K. pneumoniae isolate collection (n = 48), with a range of carbapenemases. Our flow cytometry-assisted susceptibility test (FAST) method combines rapid qualitative susceptible/non-susceptible classification and quantitative MIC measurement in a single process completed shortly after receipt of a primary isolate (54 and 158 minutes respectively). The qualitative FAST results and FAST-derived MIC (MICFAST) correspond closely with broth microdilution MIC (MICBMD, Matthew's correlation coefficient 0.887), align with the international AST standard (ISO 200776-1; 2006) and could be used for rapid determination of antimicrobial susceptibility in a wider range of Gram negative and Gram positive bacteria.

Effects of various dietary staples on esophageal carcinogenesis induced in rats by subcutaneously administered N-nitrosomethylbenzylamine.

Previous epidemiologic studies associated large differences of esophageal cancer risk with the nature of the staple diet. In this study, various cereals and dietary staples were fed to inbred BD IX rats for 7 months or longer. N-Nitrosomethylbenzylamine [(MBN) CAS:937-40-6] was given five times subcutaneously between the 45th and 58th day. The percentage of rats with tumors and the mean number of tumors per esophagus were similar when corn, wheat, commercial bird-resistant sorghum, bananas, and polished rice were fed but were strikingly lower when the basis of the diets was millet, red sorghum, brown rice, or potatoes. The number of esophageal tumors was significantly related to the dietary concentration of some minerals and vitamins. Supplementing marginally deficient corn or wheat diets with various combinations of nicotinic acid, riboflavin, zinc, magnesium, molybdenum, and selenium significantly reduced the numbers of esophageal tumors. When the feeding of protective cereals or nutrients was commenced only 150 days after MBN was given, a marked inhibitory effect on the progression of tumors was still observed.

Symptom management in patients with established renal failure managed without dialysis.

Increasing numbers of patients with chronic kidney disease Stage 5 (GFR <15ml/minute) are being managed without dialysis, either through their own preference or because dialysis is unlikely to benefit them. This growing group of patients has extensive health care needs. Their overall symptom burden is high, and symptom prevalence matches or exceeds that in other end of life populations, both with cancer and other non-cancer diagnoses. These symptoms may often go unrecognised and under-treated. Regular symptom assessment is necessary, together with pro-active management of identified symptoms. Pain can be managed using the principles of the World Health Organisation analgesic ladder. Not all opioid medications are recommended for these patients. Paracetamol, tramadol, and fentanyl are the most appropriate medications for steps 1, 2 and 3 respectively. There is limited evidence on the use of buprenorphine, oxycodone and hydromorphone. Methadone is safe but should only be prescribed by a clinician experienced in its use. Morphine and diamorphine are not recommended because of metabolite accumulation. Pruritus is also challenging to manage. The evidence for pharmacological interventions to alleviate pruritus is summarized, and a pragmatic approach to management suggested. Emollients, capsaisin cream, antihistamines, thalidomide and ondansetron may be helpful, according to the extent and pattern of pruritus. Symptoms may frequently be due to co-morbid conditions, not renal disease itself, and managing them is difficult because of the constraints on the use of medication which kidney failure imposes. Collaboration between renal and palliative specialists can help identify ways to achieve best care for these patients.

Cognitive Function in Parkinson's Disease Patients with and without Anxiety.

Research on the implications of anxiety in Parkinson's disease (PD) has been neglected despite its prevalence in nearly 50% of patients and its negative impact on quality of life. Previous reports have noted that neuropsychiatric symptoms impair cognitive performance in PD patients; however, to date, no study has directly compared PD patients with and without anxiety to examine the impact of anxiety on cognitive impairments in PD. This study compared cognitive performance across 50 PD participants with and without anxiety (17 PDA+; 33 PDA-), who underwent neurological and neuropsychological assessment. Group performance was compared across the following cognitive domains: simple attention/visuomotor processing speed, executive function (e.g., set-shifting), working memory, language, and memory/new verbal learning. Results showed that PDA+ performed significantly worse on the Digit Span forward and backward test and Part B of the Trail Making Task (TMT-B) compared to the PDA- group. There were no group differences in verbal fluency, logical memory, or TMT-A performance. In conclusion, anxiety in PD has a measurable impact on working memory and attentional set-shifting.

Anxiety is associated with freezing of gait and attentional set-shifting in Parkinson's disease: A new perspective for early intervention.

Previous research has shown that anxiety in Parkinson's disease (PD) is associated with freezing of gait (FOG), and may even contribute to the underlying mechanism. However, limited research has investigated whether PD patients with FOG (PD+FOG) have higher anxiety levels when compared directly to non-freezing PD patients (PD-NF) and moreover, how anxiety might contribute to FOG. The current study evaluated whether: (i) PD+FOG have greater anxiety compared to PD-NF, and (ii) anxiety in PD is related to attentional set-shifting, in order to better understand how anxiety might be contributing to FOG. In addition, we explored whether anxiety levels differed between those PD patients with mild FOG (PD+MildFOG) compared to PD-NF. Four hundred and sixty-one patients with PD (231 PD-NF, 180 PD+FOG, 50 PD+MildFOG) were assessed using the Freezing of Gait Questionnaire item 3 (FOG-Q3), Hospital Anxiety and Depression Scale (HADS), Digit Span Test, Logical Memory Retention Test and Trail Making Tests. Compared to PD-NF, PD+FOG had significantly greater anxiety (p<0.001). PD+MildFOG, however, demonstrated similar levels of anxiety as the PD+FOG. In all patients, the severity of anxiety symptoms was significantly correlated to their degree of self-reported FOG on FOG-Q3 (p<0.001) and TMT B-A (p=0.039). Similar results were found for depression. In conclusion, these results confirm the key role played by anxiety in FOG and also suggest that anxiety might be a promising biomarker for FOG. Future research should consider whether treating anxiety with pharmacological and/or cognitive behavioural therapies at early stages of gait impairment in PD may alleviate troublesome FOG.