Artificial intelligence-guided image acquisition on patients with implanted electrophysiological devices: results from a pivotal prospective multi-center clinical trial

Background

A novel, recently FDA-authorized software uses deep learning (DL) to provide prescriptive transthoracic echocardiography (TTE) guidance, allowing novices to acquire standard TTE views. The DL model was trained by >5,000,000 observations of the impact of probe motion on image orientation/quality. This study evaluated whether novice-acquired TTE images guided by this software were of diagnostic quality in patients with and without implanted electrophysiological (EP) devices, focusing on RV size and function, which were thought to be sensitive to EP devices. Some aspects of the study have previously been presented.

Methods

240 patients (61±16 years old, 58% male, 33% BMI >30 kg/m2, 91% with cardiac pathology) were recruited. 8 nurses without echo experience each acquired 10 view TTEs in 30 patients guided by the software. 235 of the patients were also scanned by a trained sonographer without assistance from the software. 5 Level 3 echocardiographers independently assessed the diagnostic quality of the TTEs acquired by the nurses and sonographers to evaluate the effect of EP devices on DL software performance.

Results

Nurses using the AI-guided acquisition software acquired TTEs of sufficient quality to make qualitative assessments of right ventricular (RV) size and function in greater than 80% of cases for patients with and without implanted EP devices (Table). There was no significant difference between nurse- and sonographer-acquired scans.

Conclusion

These results indicate that new DL software can guide novices to obtain TTEs that enable qualitative assessment of RV size even in the presence of implanted EP devices. The results of the comparison to sonographer-acquired exams indicate the software performance is robust to presence of pacemaker/ICD leads visible in the images (Figure).

Nurse-acquired TTE with visible ICD lead

Funding Acknowledgement

Type of funding source: Private company. Main funding source(s): Caption Health, Inc.

MiR-129 reduces CDDP resistance in gastric cancer cells by inhibiting MAPK3

OBJECTIVE

Abnormal expression of mitogen-activated protein kinase 3 (MAPK3) is related to invasion, metastasis, and drug resistance of multiple tumor cells. MiR-129 expression is associated with gastric cancer. Bioinformatics analysis showed a targeting relation between miR-129 and MAPK3. This study investigated whether miR-129 plays a role in regulating MAPK3 expression, affecting proliferation, apoptosis, and cisplatin (CDDP) resistance of gastric cancer cells.

MATERIALS AND METHODS

The dual-luciferase reporter gene assay was used to assess the targeted regulation between miR-129 and MAPK3. The expression of miR-129 and MAPK3 in CDDP-resistant cell line MGC-803/CDDP and the parental MGC-803 cells was measured. MGC-803/CDDP cells were cultured in vitro and divided into miR-NC group and miR-129 mimic group. The expression of MAPK3 and p-MAPK3 protein were detected by Western blot and the effect of CDDP treatment on cell apoptosis and proliferation was detected by flow cytometry.

RESULTS

There was a targeted regulation relation between miR-129 and MAPK3 mRNA. MiR-129 expression in MGC-803/CDDP cells was significantly lower than that in MGC-803 cells and the expression of MAPK3 mRNA and protein was significantly higher than that in MGC-803 cells. Compared with miR-NC group, the expression of MAPK3 and p-MAPK3 in MHC-803/CDDP cells in miR-129 mimic transfection group was significantly decreased, with increased cell apoptosis and reduced cell proliferation.

CONCLUSIONS

The decreased expression of miR-129 and the up-regulation of MAPK3 are associated with CDDP resistance in gastric cancer cells. Overexpression of miR-129 inhibits MAPK3 expression and cell proliferation, it induces cell apoptosis and reduces CDDP resistance.

Association of FLAIR vascular hyperintensity and acute MCA stroke outcome changes with the severity of leukoaraiosis

PURPOSE

The clinical significance of FLAIR vascular hyperintensity (FVH), a marker of collateral circulation in ischaemic stroke, remains controversial. We hypothesised that the association between FVH and outcomes varies with the severity of leukoaraiosis (LA), another marker of collaterals, and that their combined significance may vary with time.

METHODS

We included 459 consecutive patients with middle cerebral artery (MCA) stroke. Proximal and distal FVHs were distinguished based on location. LA was divided into two grades, according to Fazekas scores of 0-2 and 3-6. Symptom-to-MRI time was divided into two categories: ≤ 14 days and ≥ 15 days.

RESULTS

We found no difference in FVH proportion according to LA grade. Multivariate analysis revealed that LA and FVH status were independently associated with unfavourable outcomes (modified Rankin scale ≥ 2) in patients with symptom-to-MRI times ≤ 14 days (P = 0.008), but not in those with symptom-to-MRI times ≥15 days (P = 0.61). In the group with symptom-to-MRI times ≤14 days, patients with LA 3-6 and FVH(+) (OR, 3.044; 95% CI, 1.116-8.305) were more likely to have unfavourable clinical outcomes compared with patients with LA 0-2 and FVH(+) but not compared with those with LA 0-2 and FVH(-) or LA 3-6 and FVH(-). In addition, FVH location did not influence the relationship between FVH and outcomes.

CONCLUSIONS

The association between FVH and outcomes was influenced by the degree of LA in the acute but not in the subacute and chronic stages of MCA infarction. FVH predicts clinical outcomes independently only in those with more extensive LA.

[Relations between high risk sexual behavior and HIV infection among men who have sex with men in ways of meeting male partners]

Objective: To understand the relations between high risk sexual behavior and HIV infection among MSM in ways of finding male partners in Ningbo. Methods: A cross-sectional study was conducted in Ningbo between April and November in 2018. Data related to socio-demographics, ways of finding male partners, adoption of gay apps and sexual behaviors were collected by snowball method. Blood samples were drawn for HIV antibody testing. Classified data was evaluated by chi-square test. Related factors on HIV infection were analyzed by multivariate logistic regression. Results: A total of 735 participants were included in this study. Ways of finding male partners would through gay apps (60.8%, 447/735), QQ/Wechat (32.3%, 237/735) and gay-places (6.9%, 51/735). Related information on high risk sexual behavior and HIV infection among gay apps users were found as: 16.8%(75) had sexual behavior once per week in the past 6 months, 41.8% (187/447) had multiple sexual partners, 12.1% (54/447) had unprotected anal intercourse in the last time, 52.3% (234/447) having had unprotected anal intercourse in the past 6 months. The overall HIV prevalence rate was 12.1%(54/447). Among the HIV cases who got infected within the two years, 68.6%(24/35) of them had used gay apps for less than two years. Results from the, multivariate logistic regression analysis showed that gay apps users were more susceptible to infected HIV than those who used the QQ/Wechat (OR=3.03, 95%CI: 1.30-7.07). Conclusions: Gay apps was popularly known among the Ningbo MSM, and was associated with the high risk sexual behaviors and HIV infection. HIV control and prevention programs should be strengthened in the MSM population who used the gay apps. Related surveillance and intervention programs for MSM, who use the gay apps, need to be further reinforced.

547 Point-of-care cardiac assessment using machine learning to guide image acquisition

Funding Acknowledgements

Bay Labs, Inc; San Francisco, CA

Background/Introduction: When used by experienced examiners, the utility of point-of-care (POC) ultrasound for assessing cardiac anatomy and function has been well established. However, in some clinical circumstances (Primary Care offices, Intensive Care Unit, some Emergency Rooms, or in remote settings) in which a rapid assessment of cardiac anatomy and dynamics can facilitate patient care, an examiner experienced at POC scanning may not be immediately available.

Purpose

To help novice users acquire clinically useful standard cardiac views using novel machine learning (ML) software.

Methods

We used an investigational device that employs ML software to provide real-time adaptive guidance of transducer position and orientation to help novice users acquire tomographic views of the heart. We tested the utility of this approach when 4 nurses with no prior training in sonography performed POC studies on 16 subjects (10 healthy, 6 with cardiac abnormalities; 9 men; body mass index normal in 6, overweight in 6, and obese in 4 subjects). Each nurse underwent didactic training and 4 hours of supervised practice using the ML program. Each nurse scanned each study subject using a scanner equipped with ML software to acquire 10 digital two-dimensional image clips, including: parasternal long axis, short axis at the aortic valve, mitral valve, and mid-left ventricle (LV), apical 2-, 4-, and 5-chamber, subcostal 4-chamber, and longitudinal views of the inferior vena cava (IVC). All video clips (n = 640) were later reviewed independently by 5 level 3-trained cardiologists who were blinded to subject, scanner, and each other"s assessments. The expert readers reviewed each set of 10 clips to determine if the following variables could be assessed qualitatively: LV size and function; right ventricular (RV) size and function; aortic, mitral and tricuspid valves; pericardial effusion; left atrial size; IVC size.

Results

The majority of expert readers concurred, independently, that the sets of images acquired by nurses using ML guidance allowed qualitative assessment of LV size and function in 98%, pericardial effusion in 98%, RV size and function in 92%, and aortic and mitral valve anatomy and dynamics in 94-97% of cases. Qualitative assessment of LA size was feasible in 95%. Images of the IVC were judged as adequate for assessment in 58%.

Conclusion

This preliminary study suggests the potential value of novel ML software by demonstrating that nurses with limited training can acquire tomographic images useful for qualitative assessment of the cardiac chambers and valves in more than 90% of the subjects examined. This approach might be useful when timely POC cardiac assessment is indicated in settings where an experienced examiner is not available. Further refinements in the guiding software are needed to improve the success rate of IVC imaging, since IVC size can be a useful indicator of volume status.

Outcomes of Systemic Lupus Erythematosus in Patients who Discontinue Hydroxychloroquine

Background

Hydroxychloroquine (HCQ) is an antimalarial drug that is recommended as a safe, daily prophylactic intervention for individuals with systemic lupus erythematosus (SLE) based on previous studies that showed an association of HCQ use with reductions in flares compared with placebo. Our study aims to determine whether the discontinuation of HCQ leads to relapse of disease and whether the duration of HCQ use impacts the success of its eventual discontinuation.

Methods

A retrospective chart review was performed on the medical records of patients diagnosed with SLE between July 1, 2006, and June 30, 2016. The data gathered included demographic factors, diagnostic symptoms, laboratory values, and SLE medications. Additionally, HCQ usage and discontinuation rates were collected as well as the timing and prevalence of flares during and after HCQ usage. Patients who were diagnosed with SLE but never used HCQ were excluded from the study. The occurrence of flares, clinical characteristics, and duration of treatment with HCQ were compared between the group that continued HCQ and the group that discontinued HCQ.

Results

Of the 509 patients who met inclusion criteria, 66.2% (n = 337) continued HCQ throughout the duration of their treatment (median duration of HCQ treatment was 8.0 years), whereas 33.8% (n = 172) did not (median duration of HCQ treatment was 1.9 years). Patients who received HCQ for less than 1 year before discontinuation (median duration of HCQ treatment was 2.5 months) were more likely to experience SLE flares compared with those who continued HCQ for more than 1 year (13.1% vs 5.7%, P = 0.019). Patients who experienced a flare while on HCQ were more likely to have arthritis, oral ulcers, leukopenia, and thrombocytopenia.

Conclusion

With over 500 patient charts reviewed, this is the largest study comparing outcomes for patients on HCQ with those who discontinued it. Patients who discontinue HCQ after being on it for less than 1 year are at greater risk for flares compared with those who take HCQ for longer than 1 year. These findings should be used to guide treatment, educate patients on the role of continued treatment with HCQ, and ultimately reduce morbidity and mortality.

P4347Automated echocardiographic quantification of left ventricular ejection fraction without volume measurements using a machine learning algorithm mimicking a human expert

Background

Echocardiographic quantification of left ventricular (LV) ejection fraction (EF) relies on either manual or automated identification of endocardial boundaries followed by standard calculation of model-based end-systolic and end-diastolic LV volumes. Recent developments in artificial intelligence resulted in computer algorithms that allow near automated detection of endocardial boundaries and measurement of LV volumes and function. However, boundary identification is still prone to errors limiting accuracy in certain patients. We hypothesized that a fully automated machine learning algorithm could be developed, which circumvents border detection and instead estimates the degree of ventricular contraction, similar to a human expert trained on tens of thousands of images.

Purpose

This study was designed to test the feasibility and accuracy of this approach.

Methods

Machine learning algorithm was developed and trained on a database of >50,000 echocardiographic studies, including multiple apical 2- and 4-chamber views, to automatically estimate LVEF (AutoEF, BayLabs). Testing was performed on an independent group of 99 unselected patients, whose automated EF values were compared to reference values obtained by averaging measurements by 3 experts using conventional volume-based technique. Inter-technique agreement was assessed using linear regression and Bland-Altman analysis of bias and limits of agreement (LOA). Consistency was assessed by mean absolute deviation (MAD) among automated estimates based on different combinations of apical views. Finally, sensitivity and specificity of detecting of EF≤35% was calculated. These metrics were compared side-by-side against the same reference standard to those obtained from conventional EF measurements by clinical readers.

Results

Automated estimation of LVEF was feasible in all 99 patients. AutoEF values showed high consistency (MAD=2.9%) and excellent agreement with the reference values: r=0.95, bias=1.0%, LOA=±11.8%, with sensitivity 0.90 and specificity 0.92 for detection of EF≤35%. This was similar to clinicians' measurements: r=0.94, bias=1.4%, LOA=±13.4%,sensitivity 0.93, specificity 0.87.

Conclusions

Machine learning algorithm for volume-independent LVEF estimation is highly feasible and similar in accuracy to conventional volume-based measurements, when compared to reference values provided by an expert panel.

Acknowledgement/Funding

Bay Labs, Inc.

Evaluation of miLINC to shorten time to treatment for rifampicin-resistant Mycobacterium tuberculosis

BACKGROUND: Achieving the 90-90-90 targets for tuberculosis (TB) will require interventions that enhance diagnosis, linkage, treatment and adherence to care. As a first step in the process, our team designed a suite of smartphone applications known as miLINC to improve time from diagnosis to treatment initiation in drug-resistant TB patients.SETTING: Three clinical locations in a large, peri-urban district in KwaZulu-Natal, South Africa.OBJECTIVE: To assess the acceptability, feasibility and impact of the miLINC mobile health applications as a solution to reducing the time from presentation to treatment initiation of rifampicin-resistant (RR) TB patients.METHODS: We used a prospective, observational quality improvement evaluation of miLINC's impact among newly diagnosed patients with RR-TB.RESULTS: A convenience sample comprising details of 6341 patients with presumptive TB were entered into miLINC. Of the 631 TB-positive sputum specimens, 41 (6.5%) were found to be RR-TB. The mean time from clinical presentation to RR-TB treatment initiation was 3 days, 21 h, 17 min.CONCLUSION: This is the first study to suggest that the time from presentation to diagnosis and to treatment initiation for patients with RR-TB can be significantly improved using an integrated approach combining technology with appropriate human resources.

The Association between FLAIR Vascular Hyperintensity and Stroke Outcome Varies with Time from Onset

BACKGROUND AND PURPOSE

FLAIR vascular hyperintensity has been recognized as a marker of collaterals in ischemic stroke, but the impact on outcome is still controversial. We hypothesized that the association between FLAIR vascular hyperintensity and outcome varies with time.

MATERIALS AND METHODS

We included 459 consecutive patients with middle cerebral artery stroke and divided them into 3 groups by symptom-to-MR imaging time (group 1, ≤7 days; group 2, 8-14 days; group 3, ≥15 days). The FLAIR vascular hyperintensity score, ranging from 0 to 3 points, was based on territory distributions of different MCA segments. The associations between FLAIR vascular hyperintensity and outcome with time were analyzed qualitatively and quantitatively.

RESULTS

No patients underwent MR imaging within 6 hours of onset. The proportion of FLAIR vascular hyperintensity (+) and severe stenosis or occlusion of MCA was not significantly dependent on time. In groups 1 and 2, FLAIR vascular hyperintensity (+) was significantly associated with larger lesions, the prevalence of flow injury, and unfavorable outcome (mRS ≥ 2). There were no such associations in group 3. Multiple logistic regressions demonstrated that FLAIR vascular hyperintensity (+) was an independent risk factor for unfavorable outcome in group 2. Infarction volume tended to increase with the increase of the distal FLAIR vascular hyperintensity score in groups 1 and 2, while declining in group 3.

CONCLUSIONS

FLAIR vascular hyperintensity is associated with unfavorable outcome within 6 hours to 14 days of onset, while the wider distribution of distal FLAIR vascular hyperintensity may be favorable beyond 14 days of onset in MCA infarction. Symptom-to-MR imaging time should be considered when assessing the prognostic value of FLAIR vascular hyperintensity.

Increased risk of aminoglycoside-induced hearing loss in MDR-TB patients with HIV coinfection

SETTING

A high proportion of individuals with multidrug-resistant tuberculosis (MDR-TB) develop permanent hearing loss due to ototoxicity caused by injectable aminoglycosides (AGs). The prevalence of AG-induced hearing loss is greatest in tuberculosis (TB) and human immunodeficiency virus (HIV) endemic countries in sub-Saharan Africa. However, whether HIV coinfection is associated with a higher incidence of AG-induced hearing loss during MDR-TB treatment is controversial.

OBJECTIVE

To evaluate the impact of HIV coinfection on AG-induced hearing loss among individuals with MDR-TB in sub-Saharan Africa.

DESIGN

This was a meta-analysis of articles published in PubMed, Embase, Scopus, Cumulative Index to Nursing and Allied Health Literature, Web of Science, Cochrane Review, and reference lists using search terms 'hearing loss', 'aminoglycoside', and 'sub-Saharan Africa'.

RESULTS

Eight studies conducted in South Africa, Botswana and Namibia and published between 2012 and 2016 were included. As the included studies were homogeneous (χ2 = 8.84, df = 7), a fixed-effects model was used. Individuals with MDR-TB and HIV coinfection had a 22% higher risk of developing AG-induced hearing loss than non-HIV-infected individuals (pooled relative risk 1.22, 95%CI 1.10-1.36) during MDR-TB treatment.

CONCLUSION

This finding is critical for TB programs with regard to the expansion of injectable-sparing regimens. Our findings lend credibility to using injectable-sparing regimens and more frequent hearing monitoring, particularly in resource-limited settings for HIV-coinfected individuals.

Beyond first-line non-anthracycline-based chemotherapy for extranodal NK/T-cell lymphoma: clinical outcome and current perspectives on salvage therapy for patients after first relapse and progression of disease

Background

Current standard treatment, including non-anthracycline-based chemotherapy and optimal combining of radiotherapy, has dramatically improved outcomes of patients with extranodal natural killer/T-cell lymphoma (ENKTL) during the last decade. This study was conducted to investigate the clinical outcome of ENKTL patients with relapsed or progressive disease after initial current standard therapy.

Patients and methods

We retrospectively reviewed patients diagnosed with ENKTL at six centers in four countries (China, France, Singapore, and South Korea) from 1997 to 2015 and analyzed 179 patients who had relapsed or progressed after initial current standard therapy.

Results

After a median follow-up of 58.6 months (range 27.9-89.2), the median second progression-free survival (PFS) was 4.1 months [95% confidence interval (CI) 3.04-5.16] and overall survival (OS) was 6.4 months (95% CI 4.36-8.51). Multivariate Cox-regression analysis revealed that elevated lactate dehydrogenase, multiple extranodal sites (≥2), and presence of B symptoms were associated with inferior OS (P < 0.05). OS and PFS were significantly different according to both prognostic index of natural killer lymphoma (PINK) and PINK-E (Epstein-Barr virus) models. Salvage chemotherapy with l-asparaginase (l-Asp)-based regimens showed a significantly better clinical benefit to response rate and PFS, although it did not lead to OS improvement. First use of l-Asp in the salvage setting and l-Asp rechallenge at least 6 months after initial treatment were the best candidates for salvage l-Asp containing chemotherapy.

Conclusions

Most patients with relapsed or refractory ENKTL had poor prognosis with short survival. Further studies are warranted to determine the optimal treatment of patients with relapsed or refractory ENKTL.

Gene-Activated Matrices for Cartilage Defect Reparation

A cartilage defect has a very limited ability to repair itself spontaneously due to the shortage of blood. Many attempts have been made to restore the integrity of cartilage in clinical and experimental studies. Recently, tissue engineering has emerged as a new protocol for lost tissue regeneration. Meanwhile, the defect-repairing environment can be improved by gene therapy methods. Gene-activated matrices (GAM) fabricated with biomaterials and plasmids fill the cartilage defects to restore the integrity of joint surface, facilitating repair cell adhesion and proliferation as well as the synthesis of extracelluar matrix. And they also serve as a local gene delivery system, inducing therapeutic agent expression at the repair site. In the present study, we fabricated two- and three-dimensional matrices from chitosan and gelatin, then added a plasmid DNA encoding transforming growth factors-β1 (TGF-β1) for cartilage defect regeneration. First, we demonstrated primary chondrocytes could maintain their biological characteristics and secrete therapeutic proteins when they were cultured onto GAM in vitro. Subsequently we inserted three-dimensional GAM into cartilage defects of rabbit knee joints. With the results of the new cartilage tissue formation, we came to the conclusion that GAM was helpful for new tissue production and this therapeutic protocol provided a cheap, simple, and effective method for cartilage defect reparation.

Can zinc protect cells from the cytotoxic effects of cobalt ions and nanoparticles derived from metal-on-metal joint arthroplasties?

Objectives

Recently, high failure rates of metal-on-metal (MOM) hip implants have raised concerns of cobalt toxicity. Adverse reactions occur to cobalt nanoparticles (CoNPs) and cobalt ions (Co²⁺) during wear of MOM hip implants, but the toxic mechanism is not clear.

Methods

To evaluate the protective effect of zinc ions (Zn²⁺), Balb/3T3 mouse fibroblast cells were pretreated with 50 μM Zn²⁺ for four hours. The cells were then exposed to different concentrations of CoNPs and Co²⁺ for four hours, 24 hours and 48 hours. The cell viabilities, reactive oxygen species (ROS) levels, and inflammatory cytokines were measured.

Results

CoNPs and Co²⁺ can induce the increase of ROS and inflammatory cytokines, such as tumour necrosis factor α (TNF-α), interleukin-1β (IL-1β) and interleukin-6 (IL-6). However, Zn pretreatment can significantly prevent cytotoxicity induced by CoNPs and Co²⁺, decrease ROS production, and decrease levels of inflammatory cytokines in Balb/3T3 mouse fibroblast cells.

Conclusion

These results suggest that Zn pretreatment can provide protection against inflammation and cytotoxicity induced by CoNPs and Co²⁺ in Balb/3T3 cells.

Cite this article: Y. Liu, H. Zhu, H. Hong, W. Wang, F. Liu. Can zinc protect cells from the cytotoxic effects of cobalt ions and nanoparticles derived from metal-on-metal joint arthroplasties? Bone Joint Res 2017;6:649–655. DOI: 10.1302/2046-3758.612.BJR-2016-0137.R2.