Is it possible to make an accurate diagnosis based only on a medical history? A pilot study on women's knee joints