Automatic Food Intake Assessment Using Camera Phones

Obesity is becoming an epidemic phenomenon in most developed countries. The fundamental cause of obesity and overweight is an energy imbalance between calories consumed and calories expended. It is essential to monitor everyday food intake for obesity prevention and management. Existing dietary assessment methods usually require manually recording and recall of food types and portions. Accuracy of the results largely relies on many uncertain factors such as user's memory, food knowledge, and portion estimations. As a result, the accuracy is often compromised. Accurate and convenient dietary assessment methods are still blank and needed in both population and research societies. In this thesis, an automatic food intake assessment method using cameras, inertial measurement units (IMUs) on smart phones was developed to help people foster a healthy life style. With this method, users use their smart phones before and after a meal to capture images or videos around the meal. The smart phone will recognize food items and calculate the volume of the food consumed and provide the results to users. The technical objective is to explore the feasibility of image based food recognition and image based volume estimation. This thesis comprises five publications that address four specific goals of this work: (1) to develop a prototype system with existing methods to review the literature methods, find their drawbacks and explore the feasibility to develop novel methods; (2) based on the prototype system, to investigate new food classification methods to improve the recognition accuracy to a field application level; (3) to design indexing methods for large-scale image database to facilitate the development of new food image recognition and retrieval algorithms; (4) to develop novel convenient and accurate food volume estimation methods using only smart phones with cameras and IMUs. A prototype system was implemented to review existing methods. Image feature detector and descriptor were developed and a nearest neighbor classifier were implemented to classify food items. A reedit card marker method was introduced for metric scale 3D reconstruction and volume calculation. To increase recognition accuracy, novel multi-view food recognition algorithms were developed to recognize regular shape food items. To further increase the accuracy and make the algorithm applicable to arbitrary food items, new food features, new classifiers were designed. The efficiency of the algorithm was increased by means of developing novel image indexing method in large-scale image database. Finally, the volume calculation was enhanced through reducing the marker and introducing IMUs. Sensor fusion technique to combine measurements from cameras and IMUs were explored to infer the metric scale of the 3D model as well as reduce noises from these sensors.

Little evidence that hepatitis C virus leads to a higher risk of mortality in the absence of cirrhosis and excess alcohol intake: the Swiss Hepatitis C Cohort Study

The objective of this study was to describe the all-cause mortality of participants in the Swiss Hepatitis C Cohort compared to the Swiss general population. Patients with hepatitis C virus (HCV) infection attending secondary and tertiary care centres in Switzerland. One thousand six hundred and forty-five patients with HCV infection were followed up for a mean of over 2 years. We calculated all-cause standardized mortality ratios (SMR) and 95% confidence intervals (CI) using age, sex and calendar year-specific Swiss all-cause mortality rates. Multivariable Poisson regression was used to model the variability of SMR by cirrhotic status, HCV genotype, infection with hepatitis B virus or HIV, injection drug use and alcohol intake. Sixty-one deaths were recorded out of 1645 participants. The crude all-cause SMR was 4.5 (95% CI: 3.5-5.8). Patients co-infected with HIV had a crude SMR of 20 (95% CI: 11.1-36.1). The SMR of 1.1 (95% CI: 0.63-2.03) for patients who were not cirrhotic, not infected with HBV or HIV, did not inject drugs, were not heavy alcohol consumers (intake, in those infected with HCV.

High protein intake reduces intrahepatocellular lipid deposition in humans

BACKGROUND: High sugar and fat intakes are known to increase intrahepatocellular lipids (IHCLs) and to cause insulin resistance. High protein intake may facilitate weight loss and improve glucose homeostasis in insulin-resistant patients, but its effects on IHCLs remain unknown. OBJECTIVE: The aim was to assess the effect of high protein intake on high-fat diet-induced IHCL accumulation and insulin sensitivity in healthy young men. DESIGN: Ten volunteers were studied in a crossover design after 4 d of either a hypercaloric high-fat (HF) diet; a hypercaloric high-fat, high-protein (HFHP) diet; or a control, isocaloric (control) diet. IHCLs were measured by (1)H-magnetic resonance spectroscopy, fasting metabolism was measured by indirect calorimetry, insulin sensitivity was measured by hyperinsulinemic-euglycemic clamp, and plasma concentrations were measured by enzyme-linked immunosorbent assay and gas chromatography-mass spectrometry; expression of key lipogenic genes was assessed in subcutaneous adipose tissue biopsy specimens. RESULTS: The HF diet increased IHCLs by 90 +/- 26% and plasma tissue-type plasminogen activator inhibitor-1 (tPAI-1) by 54 +/- 11% (P < 0.02 for both) and inhibited plasma free fatty acids by 26 +/- 11% and beta-hydroxybutyrate by 61 +/- 27% (P < 0.05 for both). The HFHP diet blunted the increase in IHCLs and normalized plasma beta-hydroxybutyrate and tPAI-1 concentrations. Insulin sensitivity was not altered, whereas the expression of sterol regulatory element-binding protein-1c and key lipogenic genes increased with the HF and HFHP diets (P < 0.02). Bile acid concentrations remained unchanged after the HF diet but increased by 50 +/- 24% after the HFHP diet (P = 0.14). CONCLUSIONS: Protein intake significantly blunts the effects of an HF diet on IHCLs and tPAI-1 through effects presumably exerted at the level of the liver. Protein-induced increases in bile acid concentrations may be involved. This trial was registered at www.clinicaltrials.gov as NCT00523562.

Effects of high selenium intake on selenium status, liver function and claw health of fattening bulls

The effects of three dietary selenium (Se) levels (0.15, 0.35 and 0.5 mg/kg dry matter (dm) and of two Se-compounds (sodium selenite and Se-yeast) on the Se-status, liver function and claw health were studied using 36 fattening bulls in a two-factorial feeding trial that lasted 16 weeks. The claw health was assessed macroscopically and microscopically. Compared to the two control diets containing 0.15 mg Se/kg dm, the intake of the diets containing 0.35 and 0.50 mg Se/kg dm significantly (P < 0.05) increased the Se-concentration in serum, hair, liver and skeletal muscle. Compared to sodium selenite the intake of Se-yeast resulted in significantly (P < 0.05) higher Se-concentration in serum, liver and hair. Concerning the claw horn quality, there was no significant difference between the different groups; the animals receiving organic Se tended to have a better histological score (P = 0.06) at the coronary band than the groups fed with sodium selenite. The serum vitamin E level decreased significantly (P < 0.05) with increasing Se-intake, which had no influence (P > 0.1) on growth and liver function parameters. With the exception of the decrease of the serum vitamin E level indicating an oxidative stress caused by a high Se-intake, no negative effects of dietary selenium exceeding recommended levels for 4 months were observed.

Management Intensive Grazing of Stockers at the Andrew Jackson Demonstration Farm (1996 – 1998)

The Andrew Jackson Demonstration Farm (AJDF) is located in central Jackson County in east central Iowa. A board of directors operates the farm for the purpose of demonstrating different production practices and management strategies. From 1996 to 1998 management intensive grazing practices and the grazing of stockers on a combination of permanent and tillable pasture have been demonstrated. Grazing strategies or practices demonstrated during these years included establishment of Eastern Gamagrass and Big Bluestem, variable density grazing, measuring forage on-offer, estimating dry matter intake, grazing corn, pasture renovation, and fencing and water systems. Production performance data were gathered for the three years stockers that were grazed. During this time the stockers averaged 121 animal days of grazing, a 1.1 head per acre stocking rate, a 1.85 pound average daily gain, and 228 pounds of gain per acre. The financial measures evaluated the value of gain on pasture and the pasture cost of the gain. The value of gain per pound was positive for 1996 and 1997 at $.58 and $.52 whereas in 1998 it was a -$.04. Pasture costs per pound of gain ranged from $.12 to $.16. Production performance is only one part of the profit picture when evaluating a stocker operation. Buysell margins are the other significant part that can greatly impact the profit potential of a summer grazing program.