Identifying predictive features of autism spectrum disorders in a clinical sample of adolescents and adults using machine learning.

Diagnosing autism spectrum disorders (ASD) is a complicated, time-consuming process which is particularly challenging in older individuals. One of the most widely used behavioral diagnostic tools is the Autism Diagnostic Observation Schedule (ADOS). Previous work using machine learning techniques suggested that ASD detection in children can be achieved with substantially fewer items than the original ADOS. Here, we expand on this work with a specific focus on adolescents and adults as assessed with the ADOS Module 4. We used a machine learning algorithm (support vector machine) to examine whether ASD detection can be improved by identifying a subset of behavioral features from the ADOS Module 4 in a routine clinical sample of N = 673 high-functioning adolescents and adults with ASD (n = 385) and individuals with suspected ASD but other best-estimate or no psychiatric diagnoses (n = 288). We identified reduced subsets of 5 behavioral features for the whole sample as well as age subgroups (adolescents vs. adults) that showed good specificity and sensitivity and reached performance close to that of the existing ADOS algorithm and the full ADOS, with no significant differences in overall performance. These results may help to improve the complicated diagnostic process of ASD by encouraging future efforts to develop novel diagnostic instruments for ASD detection based on the identified constructs as well as aiding clinicians in the difficult question of differential diagnosis.

Cutting precision in a novel aortic valve resection tool. Research in progress.

BACKGROUND: We recently demonstrated the first in-vitro cutting results of a minimal-invasive aortic valve resection tool. The current study was designed to assess the cutting accuracy of this new device improved by the implementation of a linear motor-based propulsion unit. METHODS: Native aortic valves of isolated swine hearts (valve diameter 17.8+/-0.9 mm, mean+/-S.D.) were artificially stenosed and calcified (n=7). Subsequently, valves were resected by the use of a new aortic valve resection tool. The cutting process was performed by fitting the instrument with foldable Nitinol cutting blades (diameter 15 mm) and two software-operated linear motors combined with separated manual rotation. Aortic valve area was measured pre- and postprocedure by software-guided binary area calculation. Aortic valve residue has been determined and the grade of accuracy has been assessed via calculating the average midpoint of the neoannulus. Furthermore, radial deviation of concentricity was calculated and cutting time was measured. RESULTS: Aortic valve resection was successful in all cases and nearly all leaflets (2.5+/-0.4) with a weight of 0.22+/-0.12 g were cut. Aortic valve area increased significantly (0.3+/-0.1 cm(2) vs. 1.1+/-0.2 cm(2), P<0.001) with a mean cutting time of 49.7+/-15.0 s. Mean lateral leaflet rim within the annulus was 3.2+/-3.2 mm. Cutting precision revealed a median deviation of the cutting ring from the desired position of 1.3+/-0.6 mm (y-axis) and 1.4+/-0.5 mm (x-axis). Median center deviation of the cutting ring was 2.6+/-0.8 mm. CONCLUSIONS: The present study clearly confirmed ability of an accelerated cutting of stenotic aortic valve by the aortic valve resection tool. Nearly all leaflets were cut and a small rim was left within the annulus, hence providing an ideal 'landing zone' for the new prosthesis. Nevertheless, the aortic valve resection tool should be enhanced by adding a centering mechanism, thus achieving a more precise cutting process in order to avoid secondary damage.

A new tool for the resection of aortic valves: In-vitro results for turning moments and forces using Nitinol cutting edges.

The use of minimally invasive techniques for aortic valve replacement (AVR) may be limited for severely calcified and degenerated stenotic aortic valves. A quick resection leaving a defined geometry would be advantageous. Therefore, a new minimally invasive resection tool was developed, using rotating foldable cutting edges. This report describes the first experimental in-vitro results of measuring turning moment and forces during cutting of test specimens. Nitinol cutting edges were mounted on a simplified version of the resection instrument. The instrument shaft was combined with an exchangeable gear (1:3.71 vs. 1:5.0), and an exchangeable screw thread for accurate feed motion (0.35 mm or 0.5 mm) was implemented. Furthermore, the option of an added stabilisation body (SB) to prevent strut-torsion during cutting was tested. Tests were performed upon specially designed test specimens, imitating native calcified aortic valves. Resection was successful in all 60 samples (12 samples for each of the five configurations). Mean resection time ranged from 18.7+/-1.0 s (gear 1:3.71, screw thread 0.5, with SB) to 29.3+/-4.6 s (gear 1:5, screw thread 0.35, with SB), mean maximum turning moment ranged from 2.1+/-0.2 Nm (gear 1:3.71, screw thread 0.35, with SB) to 2.8+/-0.4 (gear 1:5, screw thread 0.35, with SB), mean maximum force from 36.0+/-11.3 N (gear 1:3.71, screw thread 0.35, with SB) to 56.3+/-10.5 N (gear 1:3.71, screw thread 0.5, without SB) and mean number of required rotations from 41.3+/-2.9 (gear 1:3.71, screw thread 0.5, with SB) to 59.1+/-3.7 (gear 1:3.71, screw thread 0.35, without SB). In summary, the positive influence of the stabilisation body could be shown. Combining the right parameters, it is possible to limit maximum cutting forces to F(max)<50 N and maximum turning moments to M(max)< 3.0 N.

In vitro results of a new minimally invasive aortic valve resecting tool.

BACKGROUND: Aortic valve replacement (AVR) using extracorporeal circulation is currently the treatment of choice for symptomatic aortic stenosis. However, patients with multiple high-risk comorbid conditions may benefit from reduced ECC time by a simplified and faster resection in conjunction with quick sutureless valve implantation. METHODS: A prototype of a new minimally invasive aortic valve resection tool equipped with rotating and foldable Nitinol cutting edges was designed. Commercially available aortic valve bioprostheses were artificially calcified (group 1: moderate calcified, n=8, group 2: severely calcified, n=8). In vitro resection was performed using a 21mm cutting blade. Resection time (RT), maximum turning moment (MTM) and number of required rotations (NR) were measured. Furthermore, particle generation during the process of cutting was obtained and quantified. RESULTS: Aortic valve cutting could be obtained without any complications in all cases. Cutting process resulted in a RT of 15.5+/-3s in group 1 compared to 34.9+/-15s in group 2 (p=0.005), MTM was 3+/-0.6Nm in group 1 compared to 3.5+/-0.6Nm in group 2 (p=0.068) and NR were 30.6+/-2.3 in group 1 compared to 48.1+/-15.5 in group 2 (p=0.007). Particle generation was 1.77+/-0.17g in group 1 compared to 1.41+/-0.44g in group 2 (p=0.047). CONCLUSIONS: These first in vitro results confirm feasibility and accelerated aortic valve resection within 30s. This new concept holds promise for very fast AVR in combination with insertion of sutureless aortic valve prosthesis, targeting for ischemic times less than 10min in the open heart situation. Finally, resection and percutaneous AVR within 1min in the beating heart situation is envisioned.

NiTinol-based cutting edges for endovascular heart valve resection: first in-vitro cutting results.

Machining of shape memory alloys based on Nitinol (NiTi) creates difficulties due to its ductility and severe strain hardening. In this experiment, different cutting edges and grinding parameters were tested to optimize cutting results on NiTi-based blades intended for endovascular heart valve resection. The cutting procedure was performed using two counter-rotating circular NiTi blades of different diameter. A rotating/punching process should be performed. Different shapes (glazed, waved, and saw tooth), different grinding techniques (manual, manual grinder, and precise milling cutter) and additionally various velocities (50 and 200 rpm) were tested on specific test specimens. Cutting forces were measured and cutting quality was examined using digital microscopy. Preliminary tests with rotating blades showed superior results using cutting edges for the punching process (150 N vs. 200 N; n=7). In a second step special test specimens were tested. Maximum cutting-force was 265 N+/-20 N (mean+/-SD; n=7). Subsequently different shapes were tested at 50 and 200 rpm using the rotating/punching method regarding alternate grinding techniques. Cutting forces were 27 N+/-7.7 N for glazed blades (n=7) at 50 rpm and 18 N+/-4.7 N at 200 rpm, waved blades (n=7) required a maximum force of 18 N+/-5 N at 50 rpm and 11 N+/-3.3 N at 200 rpm, whereas saw tooth blades (n=7) needed 17 N+/-12.7 N at 50 rpm and 9 N+/-1.2 N at 200 rpm. Precise cutting quality was only seen when using glazed blades sharpened under accurate conditions with a high-speed milling cutter. Although shape memory alloys based on Nitinol are difficult to process, and well-defined grinding parameters do not exist, acceptable results can be reached using high-speed milling cutters. Best cutting quality can be observed by using glazed blades, performing a rotating/punching process at high velocities. Lower cutting forces can be observed by using other shape-types, however this leads to lower cutting quality. Therefore, further investigations on blade-machining and velocity-testing seem to be necessary to create optimal cutting results.